ISOPHASE AND PHASE TRANSITIONS OF CELLULOSE – A SHORT REVIEW

transcript

Journal SITA, 2016, 18 (2), 14-34

Michael Ioelovich

Designer Energy Ltd, 2 Bergman Str., Rehovot 7670504 (ISRAEL)

E-mail: bd895892@zahav.net.il

Abstract

Cellulose has complicated supramolecular structure consisting of nanofibrils, which are built of

ordered crystallites (CR) and low ordered non-crystalline domains (NCD) in various CR/NCD

ratios, from 4 for nanocrystalline cellulose to zero for amorphized cellulose. Moreover,

crystallites have four main allomorphs, CI, CII, CII and CIV. In this critical review isophase

temperature transitions in NCD and phase transition of cellulose crystallites, as well as

amorphous cellulose, were described and discussed. It has been shown that due to structural

heterogeneity the non-crystalline domains have three isophase temperature transitions, where the

α1 and α2 transitions are caused by the segmental mobility in dense mesomorphous and medium

packed amorphous clusters, respectively; whereas the β transition is related to the mobility of

small segments in loose packed amorphous clusters, which probably are located on the outer

surface of nanofibrils. Under the action of water and other plasticizers all three isophase

transitions are shifted to lower temperatures. Various crystalline allomorphs and amorphous

cellulose can be converted into each other as a result of phase transitions, such as

recrystallization, decrystallization, transitions between various crystalline allomorphs, etc.

Important phase transition is a sol-gel process, when cellulose is dissolved and then regenerated

from the solution. In this review mechanism of the phase transitions and their relation to

isophase transitions of cellulose was disclosed.

Keywords: Cellulose, Structure, Isophase transitions, Phase Transitions, Mechanism

INTRODUCTION

As is known, the properties of polymer materials are changed gradually in a certain

temperature range until achieving the critical temperature at with an abrupt alteration of various

characteristics is observed. This phenomenon can signify the so called isophase transition

connected with structural reorganization within the same phase, or the phase transition, when

change of the phase state occurs [1, 2]. Typical example of isophase transition is glass or α

relaxation transition caused by a development of segmental mobility, as well as β and γ

relaxation transitions, which are connected with mobility change of some units or groups in

amorphous phase. The isophase transitions appear as inflection points on temperature

dependences of volume, deformation and mechanical characteristics; jump of thermal expansion

coefficient and heat capacity; extremum of dynamic properties; but these transitions do not

change the thermodynamic characteristics of polymers.

The phase transitions include amorphization, melting (fusion), dissolving, crystallization,

recrystallization, transitions between crystalline allomorphs and some other changes, which are

accompanied by jump of volume, mechanical properties, enthalpy and other physical and

thermodynamic characteristics.

Linear polymers with a simple structure typically have one glass temperature (Tg) and one

melting point (Tm). For example, natural rubber transits from glass into viscoelastic state at cca.

200 K and melts at cca. 290 K. However, polymers with complicated structure can have multiple

isophase and phase transitions. Even such known polymer as polyethylene has three glass

transitions at 153 K (for completely amorphous domains), 240 K (for mesomorphous domains on

surface of crystals) and at 200 K (for intermediate amorphous domains) [3, 4], as well as

different melting points depending on the density and type of crystals (folded, straightened,

spherulites).

Cellulose is a renewable natural polysaccharide, which is the most abundant organic matter

on Earth [5]. This natural polymer is an inexhaustible raw material for production of paper,

fibers, films, fillers, binders, glues, explosives, drugs and others valuable materials and

substances. Since the processing and use of cellulose and cellulosic materials can be carried out

over a wide temperature range, it is important to study the isophase and phase transitions of this

polymer. Cellulose is a linear, stereoregular, semicrystalline polysaccharide composed of

anhydroglucose units (AGU) linked by chemical β-1,4-glycosidic bonds. The linear

macromolecules joined by hydrogen bonds form nanofibrils, which are built of strong crystallites

with straightened chains and weak non-crystalline domains having twisted and curved segments

[6]. Molecular chains of cellulose pass through several crystallites and non-crystalline domains

linking them by strong chemical bonds. It was also discovered that non-crystalline domains have

heterogeneous packaging and can consist of relative dense mesomorphous clusters, amorphous

clusters inside of nanofibrils and loose amorphous clusters on the surface of nanofibrils [6, 7].

Besides, crystallites of cellulose have four main allomorphs CI, CII, CIII and CIV differ by

parameters of crystalline unit cells [8-10]. Furthermore, cellulose can be of various types, such as

natural, mercerized, regenerated, microcrystalline, nanocrystalline, amorphized, etc. Therefore, it

is not surprising that such a complex polymer as cellulose has multiple isophase and phase

transitions, which are discussed in this review.

ISOPHASE TRANSITIONS OF CELLULOSE

Currently it is supposed that primary α1 glass transition of cellulose is located at high-

temperatures, above 473 K (Tg1) [11, 12]. However, direct measurement of the exact value of Tg1

is difficult due to beginning of thermal decomposition of cellulose. To evaluate Tg1, indirect

methods were used. One of these methods was determination of glass temperatures (Tg) for

plasticized cellulose samples at temperatures below temperature of thermal decomposition; then

obtained glass temperatures were extrapolated to the value corresponding to zero content of

plasticizer [13, 14], as it is shown in Fig. 1.

Figure 1: Illustration of dependence of glass transition temperature on content of plasticizer

0 10 20 30 40

Another indirect method was determination of glass temperatures (Tg) of oligosaccharides

with various degree of polymerization (DP) [15], with the subsequent extrapolation of linear

dependence Tg = (DP-1) to zero value of DP-1 (Fig. 2).

Figure 2: Illustration of dependence of glass transition temperature (Tg) on degree of

polymerization (DP) of cellulose oligomers (oligosaccharides)

The third is a calculation method based on the ratio: Tg/Tm = 0.66 [1, 16]; where the

melting point, Tm ≈ 720-770 K, was estimated from experiment of flash fusion of cellulose

crystallites by laser beam [17]. As a result of indirect evaluations, the following primary glass

temperature of cellulose was obtained: Tg1 ≈ 490-500 K (Table 1). The apparent activation

energy of this transition was above 200 kJ/mole [18], which is typical for glass transition.

Below Tg1 at the temperatures 390-410 K, the relaxation transition was discovered by

various experimental methods (Table 1): sharp decrease of dynamic modulus [11], peak of

mechanical absorption [18], sharp increase of deformation [19], jumps of thermal expansion

coefficient [19, 20] and heat capacity [21] (see for example Fig. 3, 4).

Figure 3: Dilatometry of dry film of regenerated cellulose

0 0.1 0.2 0.3 0.4 0.5 0.6

Figure 4: Temperature dependence of volume expansion coefficient for dry samples of cotton

cellulose (CC) and regenerated cellulose (RC) with different crystallinity degrees (X)

The apparent activation energy of this transition was above 100 kJ/mole, which is higher

than for β transition. Therefore, most researchers attribute this transition to secondary α2 glass

transition [12, 18-21].

The third relaxation transition was detected at the temperatures 280-300 K by methods of

thermomechanics [19], calorimetry [24] and electron paramagnetic resonance [25], linear and

volume dilatometry [19, 20, 22, 23] (see for example Fig. 3, 4). However, changes of various

properties at transition temperature were relative small. Furthermore, the energy of activation of

this transition was about 80 kJ/mole [26], which is typical for β transition.

The fourth relaxation transition was found at 180-200 K using dynamic method of

dialectical absorption at low temperatures [26-28]. It has been proven that it is ϒ transition with

activation energy of 40-50 kJ/mole and caused by the mobility of hydroxymethyl groups in non-

crystalline domains.

Table 1: Isophase relaxation transitions of cellulose

Transition temperatures, K Interpretation Methods

490-500 Primary α1 glass transition Indirect and calculation

methods [13-15]

390-410 Secondary α2 glass transition Dynamic methods [11,

18], dilatometry [19, 20],

thermomechanics [19],

calorimetry [21].

280-300 β transition of small segments Dilatometry [19, 20, 22,

23], thermomechanics 19],

EPR [25]

180-200 ϒ transition of hydroxymethyl groups Dynamic dielectric

methods [26-28]

With decrease of crystallinity degree, the intensity of α, β and ϒ transitions increases; thus all

these transitions occur in non-crystalline domains of cellulose (see for example Fig. 4).

It is important to note that in contrast to elastomers, cellulose materials even above glass

transition temperatures show a high modulus, low deformation, as well as small changes of

various properties at transition temperatures. Therefore to discover isophase temperature

transitions in this polymer, special high-precise research methods should be used.

To explain multiple isophase transitions of cellulose the structural organization of non-

crystalline domains (NCD) of this polymer should be studied in detail. However, these

investigations are hampered by the presence of highly ordered crystalline regions, as well as by

the lack of reliable methods for the study of non-crystalline components. For instance, various

variants of NMR method gave a little information on structure of NCD. So, method CP/MAS 13C

NMR showed only the presence in NCD two “amorphous” components, non-accessible and

accessible clusters [29].

To study the structure of NCD by WAXS-method, samples of amorphized cellulose are

required. These samples can be prepared by ball-milling, saponification of cellulose acetate in

non-water alkali solutions and some others methods. The typical X-ray diffractogram of

amorphous cellulose has a wide peak with maximum at 2θ ≈ 20o (Fig. 5). Using the modified

Scherer’s equation for amorphous polymers with the shape factor K = 1.8, an average size of

ordered mesomorphous clusters of NCD can be estimated [7], Dms = 1.9 nm; whereas the average

Bragg’s distance between planes in such mesomorphous cluster, dms, was 0.45 nm.

Figure 5. X-ray scattering of amorphous cellulose

Further, the function of radial distribution of electron density, F, was also calculated for

precision of the cluster structure:

where i(s) is corrected reduced intensity; s = 4π sin/.

The F-function has five maximums of electron density at distances r of 0.1; 0.45; 0.9; 1.4 and

1.9 nm. The first maximum of F-distribution at 0.1 nm is intramolecular and relates to

superposition of lengths between various atoms in AGU of the polymer. Other maximums of

electron density are intermolecular and relate to average distance dms = 0.45 nm between layers

of AGU in mesomorphous cluster, while the third and fourth maximums – to the same distance

dms increased in two and three times. The radius of electron density at 1.9 nm is correlated with

average size of the cluster Dms. As is follows from calculations, the cluster consists of about five

AGU. These clusters are serves as centers of growth of crystallites at recrystallization process

of amorphized cellulose.

dsrsssirF )sin()(2 (1)

Figure 6. Model of mesomorphous unit cell of the clusters

(Projection)

As a result of analysis, a model of mesomorphous unit cell was proposed for the cluster.

This unit cell has parameters a and b about 0.9 nm, c = 1.034 nm and -angle about 120o (Fig.

6). Only one distance between layers in the cluster is nearly constant: dms = 0.45 nm. Therefore

only one maximum at 2θ ≈ 20o appears on the X-ray diffractogram of amorphous cellulose. The

specific gravity of the mesomorphous cluster is 1.48-1.50 g/cm3, while the average specific

gravity of NCD or the amorphous cellulose is 1.45 g/cm3 (Table 2).

Information about hydrogen bonds in NCD can be obtained using the method of IR

spectroscopy in combination with deuteration. It was found that the separated area of stretching

vibrations of OH-groups is located in a wide frequency range from 3000 to 3650 cm-1; this

indicates about the presence in NCD of H-bonds with different energies. After deuteration of

cellulose, a wide area of stretching vibrations of OD-groups in NCD appeared in the range 2300-

2680 cm-1 having maximum at 2500 cm-1. Energy of H-bonds can be calculated from OD-

frequency by the equation [30]:

EH = Eo + K √Δν (2)

where Eo ≈ 8 kJ/mol is minimum base energy; K = 0.81.

As a result it was discovered that the energy of H-bonds in NCD varies from 8 to 26 kJ/mol

(Table 2).

Table 2: Structural characteristics of non-crystalline domains of cellulose*

Cluster WP d, g/cm3 V, cm3/mol EH, kJ/mol Transitions DMC 0.6 1.48-1.50 108-110 23-25 α1 glass transition MAC 0.3 1.41-1.43 113-115 16-18 α2 glass transition LAC 0.1 1.29-1.31 124-126 8-10 β transition

Bulk NCD 1 1.45 112 20 α1, α2 and β

*WP is estimated weight part of a cluster; d is specific gravity; V is molar volume

Thus, the detailed studies showed that non-crystalline domains (NCD) of cellulose have

heterogeneous structural organization, which consists of dense mesomorphous clusters (DMC),

medium packed amorphous clusters (MAC) and loose packed amorphous clusters (LAC).

Consequence of structural heterogeneity of NCD of cellulose is the presence of three isophase

transitions, where the α1 and α2 transitions are caused by the segmental mobility in dense

mesomorphous and medium packed amorphous clusters, respectively; whereas the β transition is

related to the mobility of small segments in loose packed amorphous clusters, which probably

are located on the outer surface of nanofibrils.

Plasticization of cellulose causes a decrease of the temperatures of α1, α2 and β transitions

[12]. The most widespread and active plasticizer of cellulose is water. Therefore, it is advisable

to test the effect of this plasticizer on the relaxation transitions in NCD of cellulose.

Unfortunately, water is easily volatile substance, so its plasticizing effect can be studied only at

moderate and low temperatures. As a result, it was found lowering of temperature of β transition

under effect of water [19, 31, 32]. The activation energy of the β relaxation involves an entropy

contribution that varies with the water content [32]. However, effect of water on temperatures of

α1 and α2 transitions can be detected only at a high content of the plasticizer close to saturation

point [14, 19].

The largest number of papers was devoted to the study of the influence of water on the

glass transition. It has been shown that at high relative humidity of 80 to 100% temperature of

glass transition (Tg) for moist cellulose is located below room temperature [33]. To determine the

theoretical Tg of cellulose-water system over a wide range of water content, several calculation

methods were proposed. For this purpose, Salmen and Back [34] used Kaelbe’s equation [35]:

Tg = [Tg2 X2 + (h1/h2) Tg1 X1]/[X2 + (h1/h1) X1] (3)

where T1 and T2 are glass transition temperatures of water and amorphous cellulose; X1 and X2

are mole fraction of water and polymer; while h1 and h2 are constants of water and polymer,

respectively. Paes et al. [36] succeeded obtain Tg of amorphous cellulose with a certain moisture content,

and calculate the dependence of Tg on moisture content by Couchman–Karasz equation [37]:

Tg = (W1 ΔC1Tg1 + W2 ΔC2Tg2)/(W1 ΔC1 + W2 ΔC2) (4)

where Tg1 and Tg2 are glass transition temperatures of water and amorphous cellulose (AC); W1

and W2 are weight fractions of water and AC; ΔC1 and ΔC2 are change in heat capacity of

water at Tg1 and AC at Tg2, respectively.

In our investigations, we used the Fox equation [38]:

Tg-1 = W1Tg1

-1 + W2Tg2-1 (5)

where Tg1 and Tg2 are glass transitions temperatures of water and amorphous cellulose (AC); W1

and W2 are weight fractions of water and AC, respectively.

The results showed that calculations by means of all three equations gave the equal values

of theoretical glass transition temperature for wet amorphous cellulose, assuming that glass

transition temperature of dry polymer is 493 K and water is 136 K (Fig. 7).

Figure 7. Theoretical dependence of glass transition temperature on weight part of water in

amorphous cellulose or in NCD calculated by Eq. 3-5 and experimentally obtained points [36]

0 0.1 0.2 0.3

W1, w.p.

It should be noted that really it is difficult to obtain the correct results for samples of

amorphous or amorphized cellulose because of recrystallization of such samples under action of

water, or water and temperature [36]. In fact, also for semicrystalline cellulose samples the

experimental values of Tg cannot be obtained at temperatures more than 313-323 K, because at

higher temperatures the evaporation of water from the wet cellulose is enhanced resulting in

significant errors.

Consequently, a problem arises regarding extrapolation of the experimental values of Tg

for the wet samples to the value corresponding to dry cellulose. In Figure 8, we tried to

reproduce the experimental values of Tg experimentally obtained for wet cellulose samples in

some papers [33, 34, 39, 40].

Figure 8. Experimental values of Tg of wet cellulose samples with different crystallinity indexes

(CrI) obtained in papers [33, 34, 39, 40]. The dashed curve shows an exponential extrapolation to

Tg,α1 = 220°C for cellulose sample with CrI=70%

As can be seen from Fig. 8, all the experimental Tg are obtained in the temperature range -

50 to 80°C (223-353 K), i.e. far off 220-230oC attributed to temperature of primary α1 glass

transition of dry cellulose. To reach the primary glass transition temperature of dry sample, the

researchers carried out an arbitrary exponential extrapolation of the experimental points to Tg,α1

= 220°C (e.g. as it is shown in Fig. 8), which seems unjustified.

To compare the results of various samples obtained in [33, 34, 39, 40], we calculated the

moisture content (WNCD) in non-crystalline domains (NCD) of cellulose; furthermore, we

calculated Tg,c (K) of dry cellulose from the Eq. 5 in the form:

Tg,c-1 = [Tg,s (1- WNCD)]-1 – [Tg,w (1- WNCD)/(WNCD)]-1 (5a)

where Tg,s is experimental glass transition temperature of wet sample; Tg,w = 136 K is glass

transitions temperature of water; WNCD is weight fraction of water in NCD.

The calculations showed that the real Tg,c of dry cellulose is about 408 K, i.e. it is the

temperature of secondary α2 glass transition of dry cellulose (Table 3). Using this result, we

performed also the reverse calculation Tg,s = f(WNCD), shown in Fig. 9.

0 5 10 15 20

Moisure, % (db)

CrI=80%

CrI=66%

CrI=55%

CrI=35%

Table 3: Calculated glass transition temperature (Tg,c) of dry cellulose

WNCD, w.p. Tg,s, K Tg,c, K

0.1 345 416

0.2 300 420

0.3 255 408

0.35 235 386

Average 408

As follows from the obtained results (Table 3, Fig. 9), in fact the experimental data

presented in [33, 34, 39, 40] describe the dependence of secondary glass transition temperature

on the water content, i.e. Tg,α2 = f(W), instead of Tg,α1 = f(W).

Figure 9. Experimental and calculated values of Tg,s of wet cellulose samples

Some techniques, e.g. dilatometry, allow conduct the experiments in sealed ampoules filled

with aqueous solutions of high boiling liquid such as low-molecular PEG; thus the evaporation

of water from the sample can be avoided [14]. This technique allows study both the α1 and α2

glass transitions, and also β transition, in cellulose-water system [12, 14, 19]. Various studies

have revealed that with increase the water content in NCD, temperatures of all three transitions

decrease (Fig. 7, 9, 10).

Another problem is that for cellulose samples with various crystallinity degrees, different

dependences of transition temperature on the water content are observed. To get the same results

for various cellulose materials it is advisable to use the relative humidity (RH) instead of water

content in cellulose (Fig. 11).

0 0.1 0.2 0.3 0.4

WNCD, w.p.

Figure 10. Dependences of transition temperatures on weight part of water in NCD

Figure 11. Dependences of transition temperatures in cellulose-water system on relative

humidity

Like water, also other polar liquids such as triethylphenyl-ammonium hydroxide [13],

monoethanolamin, ethylenediamine, ethylene glycol, glycerol, and some others show plasticizing

effect on NCD of cellulose and cause the decrease of all transition temperatures [12, 14].

PHASE TRANSITIONS OF CELLULOSE

As known, various cellulose samples contain non-crystalline (amorphous) phase and

different crystalline allomorphs (CI, CII, CIII and CIV), which can be converted into each other

as a result of phase transitions, such as recrystallization, decrystallization, transitions between

various crystalline allomorphs, etc. In should be noted that in contrast to synthetic crystalline

polymers, crystallites of cellulose cannot melt or crystallize from melt, since theoretical melting

point is significantly higher than the temperature of chemical decomposition [17]. Some phase

0 0.1 0.2 0.3 0.4

WNCD, w.p

1 - α1

2 - α2

3 - β

[12, 19]

0 20 40 60 80 100

transitions of cellulose can be carried out within a solid state, when NCD turn into viscoelastic

state under action of temperature and/or plasticizers, or into swollen state under action of special

swelling agents. Other phase transitions may be of sol-gel type, when cellulose is dissolved and

then regenerated from the solution.

Model of non-crystalline phase, such as amorphous cellulose (AC), can be prepared using

ball-milling of semicrystalline cellulose samples, saponification of amorphous cellulose acetate

in non-aqueous medium, regeneration cellulose from solution in SO2-amine-organic solvent

system or some others methods [36, 41]. Amorphous cellulose has unstable thermodynamic

phase state, which nevertheless can be maintained for a long time in the glassy state where the

segmental motion is kinetically hindered. However, when the ambient temperature (Ta) is above

Tg (or Tg is below Ta) it makes possible the development of cooperative segmental motion and

implementation of recrystallization process within a solid viscoelastic amorphous state.

Hatakeyama et al. [42, 43] established that the wide peak of X-ray diffractogram of dry AC

became narrower after annealing at temperature of about 400 K, which is a first sign of

recrystallization. In the presence of such plasticizer as water, the crystallization of AC is

observed at room temperature (293-300 K). Study of the effect of relative humidity (RH) at room

temperature (RT) showed that the first sign of recrystallization of AC was observed at RH of 65-

75 % [44].

Paes et al. [36] reported on starting of recrystallization process after conditioning of AC at

54% RH and RT followed by drying. On the other hand, it has been found that recrystallization

of AC is caused by moisture, but not by drying [44]. In all mentioned cases the experiments were

carried out in such conditions, where the ambient temperature, Ta, exceeded the temperature of

secondary α2 glass transition, whereas the temperature of primary α1 glass transition was higher

than Ta (Fig. 11).

To achieve the temperature of primary α1 glass transition at RT, the relative humidity

should be above 80%. At these conditions, the recrystallization process is carried out more

intensive and leads to perceptible increase in crystallinity of cellulose [44]. Liquid water causes

intensive recrystallization of AC in a wide temperature range, above melting point [45]. After

wetting, the completely amorphized cellulose recrystallizes in CII crystallites; while for

crystallization in the CI allomorph, the amorphized sample must contain residual CI crystallites

[46, 47].

Two-stage crystallization process of AC takes place (Fig. 12) after treatment also with

other plasticizers such as ethylene glycol (EG) and glycerol (GL) [45].

Figure 12. Dependence of crystallinity (X) of ball-milled cellulose on heating temperature in

medium of EG (1) and GL (2)

When AC is heated in the medium of liquid plasticizer above temperature of secondary α2

glass transition in this medium, the formation of crystalline nuclei with lateral size (Ln) of 1.5-

1.8 nm is observed. On the second stage, after heating of the sample in the liquid medium above

temperature of primary α1 glass transition in this medium, an intensive recrystallization of AC

on the nuclei occurs into crystallites with lateral size (Lcr) of 3.5-4.0 nm, which is accompanied

by an increase in crystallinity (X) to 50-56% (Table 4). Non-polar or weakly polar liquids, such

as low-molecular polyethylene glycol (PEG), lower alcohols, esters, ethers and hydrocarbons do

not plasticize and do not recrystallize AC.

Table 4: Recrystallization of AC in various liquid media*

Medium Tg,α2, K Tn, K Ln , nm Tg,α1, K Tcr, K Lcr , nm X, % PEG 400 410 1.5 (493) - - 20

GL 380 400 1.8 460 470 4.0 51

EG 330 350 1.6 375 400 3.5 50

H2O 240 <273 - 260 <273 3.5 56

Mechanism of recrystallization of the completely amorphized cellulose under action of

water is, probably, in: hydration of the sample, transformation of the plasticized amorphous

phase into viscoelastic state with developed cooperative segmental motion; growth of hydrated

crystallites on nuclei - mesomorphous clusters of AC; and finally the conversion of hydrated

crystalline into CII crystallites after drying. In the case of recrystallization of partially

amorphized natural cellulose under action of water and other plasticizers, energetically more

advantageous is the crystallization of viscoelastic NCD on the existing CI-matrix with forming

of CI crystallites.

Important type of phase transition is a sol-gel process, which occurs in the production of

regenerated fibers and films using various solvents such as cuprammonium (Cupra process),

NaOH/CS2 (Viscose process), N-Methylmorpholine N-oxide (Lyocell process), etc. As a result

of regeneration, a cellulose solution turns into amorphous or mesomorphous hydrogel, which

after stretching and drying at increased temperatures transforms into regenerated cellulose with

crystallinity of 35-40% and containing small CII-crystallites. The hydrogel isolated from N,N-

dimethylacetamide/LiCl solvent system followed by stretching was converted into a dry film

with low crystallinity less than 20%, having predominantly a nematic mesomorphous structure

[41]. Further, the nematic cellulose was mercerized with 7% NaOH, treated with 50% EDA or

heated in water at 453 K; as a result, the CII, CIII and CIV allomorphs respectively were

obtained.

As known, crystallites of various cellulose samples can be in different crystalline

allomorphs. Crystallites of natural celluloses have the allomorph type CI. Furthermore it was

found that crystalline unit cell of CI can be in two distinct crystalline forms: triclinic CIα of P1-

space group and monoclinic CIβ of P21-space group; where CIα form is characteristic for algae

and bacterial celluloses, while more stable CIβ form is dominant in higher plants and tunicate [6,

8-10]. Additional crystalline allomorphs: CII, CIII1, CIII2, CIV1 and CIV2, have been identified,

which are attributed to structural-modified celluloses. Various allomorphs have different

parameters and shape of crystalline unit cells.

Different crystalline allomorphs can be converted into each other under certain conditions. It is

well known how to convert one crystalline allomorph of cellulose to another. For example,

cellulose samples containing CII-crystallites can be obtained by mercerization, i.e. treatment of

natural cellulose with 4-5 M sodium hydroxide at RT, or by regeneration from cellulose

solutions. Cellulose samples with CIII1 and CIII2 crystalline allomorph are derived from samples

of CI or CII, respectively, by treatment with liquid ammonia, primary amines or

ethylenediamine. Samples with CIV1 and CIV2 crystallites can be prepared usually by heating of

other allomorphs in hot glycerol at 533 K. When processing cellulose in the hot glycerol, this

liquid medium isolates the sample from the air atmosphere, thereby preventing the development

of thermo-oxidative destruction.

Due to existence of distinct crystalline polymorphs and amorphous cellulose, various

studies were performed in order to estimate the phase stability of different allomorphs. The

amorphous phase state is regarded as a labile, because the amorphous cellulose can easily

recrystallize in any crystalline allomorph under certain conditions [44-47]. This conclusion is

also confirmed by the results of thermodynamic and thermochemical investigations [48-50]. In

the case of crystalline polymorphs of cellulose, the problem regarding the relative stability of the

phase state is not completely resolved and remains open. Study of thermodynamic and

thermochemical characteristics gave reason to believe that phase state of CII is more stable than

CI [48-50]. However, in another study it was concluded that the stability of phase state of

various crystalline allomorphs cannot be estimated with a reasonable degree of certainty due to

uncertainties in the measured characteristics [50].

As is known, the completely crystalline cellulose is absent. Therefore, to study the phase

state of a crystalline allomorph, it requires the use of several samples with different crystallinity

degrees. Unfortunately, the existing methods for determining the crystallinity of cellulose such

as XRD and 13C-NMR, give unreliable indexes of crystallinity (CrI) [6], which cannot be used in

accurate thermodynamic calculations. To avoid the use of uncertain CrI, a precise

thermochemical method was proposed to determine the actual degree of crystallinity of cellulose

samples. This thermochemical method is based on the measurement of standard enthalpy of

cellulose wetting with water (ΔwHo) [6, 51]. The actual degree of crystallinity (X) of cellulose

was calculated by the equation:

X= 1- (ΔwHo/ΔwHoam) (6)

where ΔwHoam = - 27.16 kJ/mol is enthalpy of wetting of amorphous cellulose.

Various cellulose samples were studied. Samples with CIβ allomorph were: chemical Kraft

pulp (KP), cotton cellulose (CC), Avicel MCC PH-101 (MCC-1) and PH-301 (MCC-2)

Amorphization of cellulose was carried out by ball-milling of Kraft pulp with ceramic balls for 2

(AC-1) and 8 h (AC-2). Samples of KP and CC were treated with 20% NaOH at room

temperature overnight to obtain mercerized samples with CII allomorph - KPM and CCM,

respectively. Besides, fibers of regenerated cellulose (RC) with CII allomorph were used. To

prepare CIII from KP (KPA) and CC (CCA), the cellulose samples were treated with anhydrous

liquid ammonia at 240 K for 3 h. Samples with CIV allomorph – KPAG and CCAG, were

prepared by treatment of KPA and CCA samples in glycerol at 533 K for 30 min.

The combustion reaction of one glucopyranose unit of cellulose is written as:

C6H10O5 (s) + O2 (g) = 6CO2 (g) + 5H2O (l) + ΔcHo (7)

Based on this reaction, the standard enthalpy of formation of cellulose can be calculated by the

equation:

ΔfHo = 6ΔfH

o (CO2, g) + 5ΔfHo (H2O, l) – ΔcH

where standard enthalpies of formation of carbon dioxide and water are:

ΔfHo (CO2, g) = -393.51 kJ/mol and ΔfH

o(H2O, l) = -285.83 kJ/mol.

Gibbs free energy of formation was calculated, as follows:

ΔfGo = ΔfH

o –To(So - ∑𝑆𝑖) (9)

where ∑𝑆𝑖 is sum of standard entropies of carbon atoms (graphite), molecules of H2 and O2

needed for forming one glucopyranose unit of cellulose; To = 298.15 K.

The thermodynamic characteristics of cellulose samples having different structural

characteristics are given in Table 5.

Table 5: Standard thermodynamic characteristics of cellulose samples [52]

Abbreviation Allomorph X -ΔcHo, kJ/mol -ΔfHo, kJ/mol -ΔfGo, kJ/mol MCC-1 CI 0.72±0.02 2821.0±2.0 969.2±2.0 667.0±2.3

MCC-2 CI 0.75±0.02 2819.8±1.8 970.4±1.8 668.1±2.1

CC CI 0.70±0.02 2821.2±1.7 969.0±1.7 666.8±2.0

KP CI 0.65±0.02 2823.6±2.2 966.6±2.2 664.6±2.5

AC-1 CI 0.52±0.02 2828.7±2.3 961.5±2.3 659.9±2.6

AC-2 CI 0.28±0.02 2837.4±2.1 952.8±2.1 652.0 ±2.4

CCM CII 0.55±0.02 2823.3±1.8 966.9±1.8 664.7±2.1

KPM CII 0.53±0.02 2824.2±2.3 966.3±2.3 664.4±2.6

RC CII 0.40±0.02 2830.0±1.7 960.2±1.7 659.0±2.0

CCA CIII 0.37±0.02 2836.4±2.3 953.8±2.3 652.7±2.6

KPA CIII 0.35±0.02 2837.0±2.0 953.2±2.0 652.2±2.3

CCAG CIV 0.60±0.02 2825.0±2.1 965.2±2.1 662.4±2.4

KPAG CIV 0.57±0.02 2826.1±1.9 964.1±1.9 662.2±2.2

As can be seen from Figures 13 and 14, with the decrease of crystallinity degree all linear

dependences ΔfHo = f(X) and ΔfG

o = f(X) converge at one common point, ΔfHoam = -942.4

kJ/mol and ΔfGoam

= -642.6 kJ/mol, corresponding to X = 0. This evidences that the non-

crystalline (amorphous) phase in different cellulose samples has identical thermodynamic

characteristics. On the other hand, the linear extrapolation of these dependences to the values

corresponding to X = 1 gives the enthalpy and free energy of formation of different crystalline

allomorphs, ΔfHocr and ΔfG

ocr (Table 6). Furthermore, melting enthalpy (ΔHo

m) of the cellulose

crystallites was calculated:

ΔHom = ΔfH

oam - ΔfH

ocr (10)

Figure 13. Dependence of enthalpy of cellulose formation on degree of crystallinity

0 0.2 0.4 0.6 0.8

Figure 14. Dependence of Gibbs energy of cellulose formation on degree of crystallinity

Table 6: Thermodynamic characteristics of various allomorphs [52]

Allomorph -ΔfHo, kJ/mol -ΔfGo, kJ/mol ΔHom, kJ/mol

CI 979.6 676.4 37.2

CII 986.9 683.7 44.5

CIII 973.2 670.0 30.8

CIV 980.4 677.2 38.0

CA* 942.4 642.6 0

*CA is cellulose in noncrystalline (amorphous) phase state

Based on obtained thermodynamic characteristics, as well as thermochemical and

physicochemical properties [44-50] it can be concluded that the relative thermodynamic stability

of the various allomorphs decreases in the following order:

CII > CIV ≥ CI > CIII > CA (11)

On the other hand, the relative reactivity of the different allomorphs decreases in the reverse

order:

CA > CIII > CI ≥ CIV > CII (12)

Increased phase stability of the CII crystallites leads to some problems at the cellulose

application. It is known that after transformation of crystalline structure CI of natural cellulose

into CII, e.g. by mercerization, a decrease in solubility and reactivity of the sample is observed

[53]. In particular, transparence and filterability of solutions of CII-samples declines due to

forming of gel-particles. The presence of CII crystallites is a main factor that causes the low

reactivity of cellulose at acetylation, nitration and forming of viscose [54, 55]. On the other

hand, the transformation of crystalline structure CI of chemical pulp to labile CIII1 allomorph

increases the reactivity of cellulose [55].

0 0.2 0.4 0.6 0.8

It should be noted that the found order of the stability or reactivity is valid only for

macrocrystals or for crystallites with the close sizes. If the sizes of nanocrystallites of two

allomorphs are very different, this order can be changed due to contribution of Gibbs-Thomson

free surface energy (ΔGos) to free energy of conversion (ΔGo

c) of nanocrystallites of allomorph 1

into crystallites of allomorph 2:

ΔGoc = ΔfG

o2 - ΔfG

o1 + ΔGo

s (13)

ΔGos = 2σ1,2 V(L1 - L2)/L1L2 (13a)

where L1 and L2 are lateral sizes of rod-like nanocrystallites for initial and converted allomorph; σ1,2 is interface energy between allomorphs 1 and 2; V is molar volume; while experimental

coefficient K= 2σ1,2 V ≈ 100 (kJ x nm/mol) if L is in nm.

For example, consider the conversion of CII into CIV. If crystallites of CII and CIV have

the same size, then ΔGos =0, and the conversion order is determined by ΔfG

o values of Table 6,

i.e. free energy of conversion will be:

ΔGoc = ΔfG

o (CIV) - ΔfGo (CII) = 6.5 (kJ/mol)

Since ΔGoc > 0, the conversion of CII to CIV cannot be implemented. Indeed, the

mercerized CII cellulose cannot be converted into CIV because the both CII and CIV crystallites

have the similar lateral size of 5.6-6.0 nm. This result corresponds to the order of stability (11),

wherein CII is more stable than CIV.

On the other hand, if lateral size of CII-crystallites (LCII = 4 nm) is less than the lateral size

of CIV-crystallites (LCII = 6 nm), then order of stability will be reverse due to contribution of

ΔGos = - 8.3 kJ/mol, and namely:

ΔGoc = ΔfG

o (CIV) - ΔfGo (CII) + ΔGo

s = 6.5 - 8.3 = - 1.8 (kJ/mol)

In this case ΔGoc < 0. Thus, the conversion of small CII-crystallites into larger CIV-crystallites is

possible, which is confirmed by experiment.

General scheme of transformation of various crystalline allomorphs of cellulose is shown

on the Figure 15 [6].

Figure 15. Scheme of phase transitions between various crystalline allomorphs of cellulose

CIV1 CIII1

CIII2 CIV2

Some allomorphs, such as CI, CIII1 and CIV1, as well as CII, CIII2 and CIV2 can be

converted into each other reversible, while conversions CI into CII or CI into CIV are

irreversible. An example of reversible phase transition is transformation of CI into CIIII after

treatment by liquid ammonia and reverse transformation of CIII1 into CI after boiling in water or

by high-temperature treatment above Tg,α1 [56].

To form CIII1 allomorph, molecules of ammonia penetrate into crystalline lattice of CI and

create crystalline ammonia-cellulose complex, which turns into CIII1 while drying. Mechanism

of reverse transformation of CIII1 into CI consists in transition of NCD of CIII1 into viscoelastic

state at heating in the plasticizing medium; crystallization of viscoelastic NCD on residual CI-

nuclei with forming of stable CI crystallites, which is accompanied by breaking of small unstable

crystallites of CIII1.

Mechanism of irreversible transformation of CI into CII during mercerization consists in

penetration of hydroxyl anions into crystalline lattice of CI with creation of crystalline alkali-

cellulose, which turns into crystallites of cellulose hydrate after washing, and then into CII

crystallites after drying [52].

Unfortunately, the scheme 15 is simplified and only partially describes the real phase

transformations. Especially, it concerns the formation mechanism of CIV allomorph. For

example, small CII-crystallites (lateral size Lcr = 4.0 nm) of regenerated cellulose (RC) actually

are transformed into CIV after treatment in a hot glycerol at the conventional conditions, 533 K

for 30 min (Fig. 16). However, at the same treatment conditions, larger CII-crystallites (Lcr = 5.7

nm) of mercerized cotton cellulose (CC) are not turned into CIV allomorph (Fig. 17).

Another example is related to phase transition of CI to CIV after treatment in a hot glycerol

at the conventional conditions. As it was shown in the paper [56], CI-crystallites of natural

cellulose samples (cotton cellulose, wood pulp, etc.) cannot be transformed directly into CIV,

which is contrary to scheme 15. Nevertheless, after partial decrystallization of natural cellulose

samples e.g. by ball-milling, the remained small crystallites of CI (Lcr = 3.4 nm) are easily

converted into CIV allomorph after treatment in hot glycerol (Fig. 18). Thus, the experimental

data confirm the scheme 15 concerning formation of CIV allomorph, but only for small

crystallites having lateral size Lcr <4.3 nm.

Figure 17. X-ray diffractograms of

mercerized CC (Lcr = 5.7 nm) (1) and this

cellulose treated in hot glycerol (2)

Figure 16. X-ray diffractograms of

RC (Lcr = 4.0 nm) (1) and this

cellulose treated in hot glycerol (2)

Figure 18. X-ray diffractograms of ball-milled cellulose I (Lcr = 3.4 nm) (1) and this cellulose

treated in hot glycerol (2)

After study of crystalline structure of various cellulose samples we found that the main

structural factor responsible for transformation of any of three crystalline allomorphs, CI, CII and

CIII into CIV after treatment in the hot glycerol is a lateral size of crystallites (Lcr). As can be

seen from Table 7 [45], the conversion of any crystalline allomorph into CIV-structure at 533 K is

possible, when the lateral size of crystallites is ≤ 4.3 nm; whereas the larger crystallites cannot be

transformed directly into the CIV allomorph.

Table 7: Structural characteristics of cellulose samples and their ability to conversion into CIV*

Cellulose CRA Lcr , nm CIV Index Cotton cellulose (CC) CI 8 0

Partially decrystallized CC CI 3.6 1

Mercerized CC CII 5.7 0

Kraft pulp (KP) CI 6.7 0

Partially decrystallized KP CI 3.4 1

Mercerized KP CII 5.6 0

KP treated by liquid NH3 CIII 3.3 1

Regenerated cellulose fibers CII 4.0 1

Regenerated cellulose film CII 4.3 1

*Note: If the CIV index is zero it means the absence of phase transformation; If CIV index is one, it means

complete transformation into CIV.

Prediction of phase transformation to form of CIV allomorph was shown above by

contribution of free surface energy to free energy of conversion (see eq. 13). Quantitative

calculations of this phase transition can be performed by means of thermodynamic equation of

Gibbs and Thomson, which describes the dependence of melting point (Tm) of small rod-like

crystallites on their lateral size, Lcr [57]:

Tm = Tm,o – (2σc-a V Tm,o/ΔHm Lcr) (14)

where Tm,o is equilibrium melting point of large crystals; ΔHm is melting enthalpy; V is molar

volume of crystals; and σc-a is crystal – melt interface energy.

Thermal analysis of various cellulose samples in inert atmosphere revealed that in the

temperature range of 520 to 620 K the decay of crystalline structure occurs until complete

amorphization, which triggers the subsequent chemical decomposition [58]. Moreover, it was

found that an average temperature of thermal amorphization (Ta) is inversely proportional to

lateral size of crystallites (Fig. 19):

Ta = Ta,o – n/Lcr (15)

where coefficient n = 893 K, if Lcr is expressed in nm.

Figure 19. Dependence of average temperature of thermal amorphization on lateral size of

crystallites

The linear extrapolation of dependence Ta= f(Lcr-1) to Lcr

-1 = 0, gives value of Ta,o ≈ 740 K,

which is close to theoretical melting point of large crystals of cellulose [17]. The experimentally

obtained Eq. (15) is an analogue of theoretical Eq. (14), where Ta,o = Tm,o ≈ 740 K; and

coefficient n = 2σc-a V Tm,o/ΔHm.

As is follows from investigations, crystallites of CI, CII or CIII with lateral size Lcr > 4.3

nm (Lcr-1 < 0.23 nm-1) do not decrystallize at the treatment temperature of 533 K, and hence they

cannot be transformed into CIV after treatment in hot glycerol (Fig. 19). On the other hand, a

high-temperature treatment of cellulose having small crystallites with lateral size of Lcr ≤ 4.3 nm

(Lcr-1 > 0.23 nm-1) causes thermal amorphization (melting) of these crystallites. Then, an unstable

viscoelastic amorphous phase recrystallizes into CIV-crystallites with lateral size of Lcr ≈ 6.0 nm

(Lcr-1 = 0.17 nm-1), which remain intact at the processing temperature of 533 K.

FINAL REMARKS

Cellulose is a renewable natural polysaccharide, which is the most abundant organic matter on

Earth. This natural polymer is an inexhaustible raw material for production of paper, fibers,

films, fillers, binders, glues, explosives, drugs and others valuable materials and substances.

Since the processing and use of cellulose and cellulosic materials can be carried out over a wide

temperature range, it is important to study the isophase and phase transitions of this polymer.

Cellulose has complicated supramolecular structure consisting of nanofibrils, which are built of

ordered crystallites (CR) and low ordered non-crystalline domains (NCD) in various CR/NCD

ratios, from 4 for nanocrystalline cellulose to zero for amorphized cellulose. Moreover,

crystallites have four main allomorphs, CI, CII, CII and CIV.

It has been shown that due to structural heterogeneity the non-crystalline domains have

three isophase temperature transitions, where the α1 and α2 transitions are caused by the

occurrence of segmental mobility in dense mesomorphous and medium packed amorphous

clusters, respectively; whereas the β transition is related to the mobility of small segments in

loose packed amorphous clusters, which probably are located on the outer surface of nanofibrils.

Under the action of water and other plasticizers all three isophase transitions are shifted to lower

temperatures. Various crystalline allomorphs and amorphous cellulose can be converted into

each other as a result of phase transitions, such as recrystallization, decrystallization, transitions

between various crystalline allomorphs, etc. In this review mechanism of the phase transitions

and their relation to isophase transitions of cellulose was disclosed.

Amorphous cellulose has unstable thermodynamic phase state, which nevertheless can be

maintained for a long time in the glassy state where the segmental motion is kinetically hindered.

However, when the ambient temperature (Ta) is above Tg (or Tg is below Ta) it makes possible

the development of cooperative segmental motion and implementation of recrystallization

process within a solid viscoelastic amorphous state. Important phase transition is a sol-gel

process, when cellulose is dissolved and then regenerated from the solution.

Different crystalline allomorphs can be converted into each other under certain conditions.

Mechanism of phase transitions of these allomorphs was described. In particular, the mechanism

of CIV formation consists in thermal amorphization (melting) of small crystallites (Lcr ≤ 4.3 nm)

at the treatment temperature of 533 K in glycerol, and the following recrystallization of unstable

viscoelastic amorphous phase into relative large CIV crystallites.

This review addresses only main basic research. The complexity and diversity of

mechanisms of isophase and phase transitions in cellulose requires further more detailed

investigations.

REFERENCES

[1] R.F. Boyer. Transition and Relaxation in Polymers. Wiley & Sons, New York, 1967.

[2] M. Shen, A. Eisenberg. Glass transitions in polymers. Rubber Chem. Technol., 43, 95 (1970).

[3] S, Fakirov, B. Krasteva. On the glass transition temperature of polyethylene as revealed by

micro-hardness measurements. J. Macromol. Sci., B39(2), 297-301 (2000).

[4] G. Bartenev, R. Aliguliev, D. Chiteeva. Relaxation transitions in polyethylene. High-Mol.

Compounds, 23(9), 2003-2011 (1981).

[5] D. Klemm, B. Heublein, H.-P. Fink, A. Bohn. Cellulose: fascinating biopolymer and

sustainable raw material. Angew. Chem., 44, 2 - 37 (2005).

[6] M. Ioelovich. Cellulose Nanostructured Natural Polymer. LAP, Saarbrücken, 2014.

[7] M. Ioelovich, M. Laka. Structural analysis of disordered cellulose. SITA, 4, 90-93 (2002).

[8] A. O'Sullivan. Cellulose: the structure slowly unravels. Cellulose, 4, 173-207 (1997).

[9] H. Krässig. Cellulose: Structure, Accessibility and Reactivity. Gordon and Breach Publ.,

Amsterdam, 1993.

[10] P. Zugenmaier. Crystalline Cellulose and Cellulose Derivatives: Characterization and

Structures. Springer-Verlag, Berlin/Heidelberg, 2008.

[11] E. Back, E. Didriksson. Four secondary and glass transition temperature of cellulose

evaluated by sonic pulse technique. Svensk Papperstidn., 72(21), 687-694 (1969).

[12] I. Kaimin, V. Karlivan and M. Ioelovich. Temperature transitions of cellulose and their

dependence on low molecular substances. News Latvian Acad. Sci., 8, 112-123 (1979).

[13] V.A. Kargin, P.V. Kozlov, W. Naj-Tchan. On the glass transition temperature in cellulose.

Proceed. Acad. Sci., 130(2), 356-358 (1960).

[14] M. Ioelovich, I. Kaimin. Study of temperature transitions of cellulose in liquid media.

High-Mol. Compounds, 21B (8), 621-625 (1979).

[15] E. Alftha, A. Deruvo, W. Brown. Glass transition temperature of oligosaccharides.

Polymer, 14, 329-330 (1973).

[16] D.W. Van Krevelen. Properties of Polymers. Correlation with Chemical Structure, Elsevier

Publ., Amsterdam, London, New York, 1972.

[17] S. Nordin, J. Nyren, E. Back. Note on molten cellulose produced in a laser beam. Svensk

Papperstidn., 76(16), 609-610 (1973).

[18] S. Manabe, M. Iwata, K. Kamide. Dynamic mechanical absorptions observed for

regenerated cellulose solids in the temperature-range from 280 K to 600 K. Polym. J., 18

(1), 1-14 (1986).

[19] I. Kaimin, M. Ioelovich. Effect of moisture on transition temperatures of cellulose. High-

Mol. Compounds, B15 (10), 764-767 (1973).

[20] C. Klason, J. Kubat. Thermal transition in cellulose. Svensk Papperstidn, 79(15), 494-500

(1976).

[21] E. Fainberg, N. Michailov. Study of temperature dependence of heat capacity of cellulose

fibers. High-Mol. Compounds, 9(4), 920-926 (1967).

[22] M. Ramiah, D.A. Goring. The thermal expansion of cellulose, hemicellulose and lignin. J.

Polym. Sci., 11C, 27-48 (1965).

[23] K. Aziz, H.G. Shinouda. Acid hydrolysis of cotton cellulose and its thermal transition. Cell.

Chem. Technol., 7(4), 465-478 (1973).

[24] A.N. Mochalov, T.B. Chlustova, M.Y. Ioelovich, I.F. Kaimin. Effect of crystallinity degree

on heat capacity of cellulose. Wood Chem., 4, 66-68 (1982).

[25] O. Plotnikov, A. Michailov, E. Rayavee. Study of extremely slow molecular motions in

cellulose by EPR method. High-Mol. Compounds, 19(11), 2528-2537 (1977).

[26] T. Borisova, G. Pertopavlovsky, N. Kotelnikova. Study of temperature transitions in

cellulose by dielectric method. High-Mol. Compounds, 21(9), 2031-2037 (1979).

[27] G. Michailov, A. Artuhov, T. Borisova. On peculiarities of relaxation of hydroxyl groups in

cellulose at low temperatures. High-Mol. Compounds, 9B (2), 138-141 (1967).

[28] Yu. Zelenev, V. Glazkov. Relaxation processes in cellulose and its derivatives. High-Mol.

Compounds, 14(1), 16-22 (1972).

[29] K. Wickholm1, E.L. Hult, P.T. Larsson, T. Iversen, H.Lennholm. Quantification of

cellulose forms in complex cellulose materials: a chemometric model. Cellulose, 8, 139–

148 (2001).

[30] R.G. Zhbankov, P.V. Kozlov. Physics of Cellulose. Science and Technique, Minsk, 1983.

[31] R.A. Stratton. Dependence of the viscoelastic properties of cellulose on water content. J

Polym. Sci. Polym Chem., 11, 535–544 (1973).

[32] H. Montes, K. Mazeau, J.Y. Cavaille. Secondary mechanical relaxations in amorphous

cellulose. Macromolecules, 30, 6977- 6984 (1997).

[33] L. Szczesniak, A. Rachocki, J. Tritt-Goc. Glass transition temperature and thermal

decomposition of cellulose powder. Cellulose 15, 445–451 (2008).

[34] N.L. Salmen, E.L. Back. The influence of water on the glass phase transition temperature

of cellulose. TAPPI, 60(12), 137-140 (1977).

[35] K. Kaelbe. Physical Chemistry of Adhesion. Wiley-Interscience, New York, 1971.

[36] S.S. Paes, Sh. Sun, W. Mac Naughtan, R. Ibbett, J. Ganster, T.J. Foster, J.R. Mitchell. The

glass transition and crystallization of ball milled cellulose. Cellulose, 17, 693-709 (2010).

[37] P.R. Couchman, F.E. Karasz. A classical thermodynamic discussion of the effect of

composition on glass-transition temperatures. Macromolecules, 11, 117-119 (1978).

[38] T.G. Fox. Influence of diluent and copolymer composition on the glass temperature of a

polymer system. Bull. Am. Phys. Soc., 1, 123-125 (1956).

[39] H. Batzer, U.T. Kreibich. Influence of water on thermal transitions in natural polymers and

synthetic polyamides. Polym. Bulletin, 5(11), 585-590 (1981).

[40] Y. Ogiwara, H. Kubota, S. Hayashi, N. Mitomo. Temperature dependency of bound water

of cellulose studied by a high-resolution NMR spectrometer. J. Appl. Polym. Sci., 14(2),

303-309 (1970).

[41] E. Togawa, T. Kondo. Unique structural characteristics of nematic ordered cellulose -

stability in water and its facile transformation. J. Polym. Sci. Polym. Phys., 45, 2850-2859

(2007).

[42] H. Hatakeyama, T. Hatakeyama, J. Nakano. Studies on hydrogen-bond formation of

amorphous cellulose. Cellul. Chem. Technol., 8, 495-507 (1974).

[43] T. Hatakeyama, H. Hatakeyama. Thermal Properties of Green Polymers and

Biocomposites. Kluwer Acad. Publ., NewYork, 2004.

[44] D.F. Caulfield, R.A. Steffes. Water-induced recrystallization of cellulose. TAPPI, 52,

1361- 1366 (1969).

[45] M.Ya. Ioelovich, I.F. Kaimin, G.P. Veveris. The crystallization of amorphized cellulose. J.

Polym. Sci., 24(6), 1382-1387 (1982).

[46] I.L. Wadehra, R.S.J. Manley. Recrystallization of amorphous cellulose. J. Appl. Polym.

Sci., 9(7), 2627-2630 (1965).

[47] P. BhamaIyer, S. Sreenivasan, P.K. Chidambareswaran, N.B. Patil Recrystallization of

cellulose. Textile Res. J., 56, 509-511 (1986).

[48] M. Ioelovich. Thermochemistry of cellulose dissolution process. Wood Chem.,

5, 111-112 (1985).

[49] M.Ya. Ioelovich, N.A. Cemeris, M.M. Cemeris. Effect of cellulose modifying on enthalpy

of formation. Wood Chem., 4, 15–18 (1988).

[50] R.N. Goldberg , J. Schliesser , A. Mittal, et al. A thermodynamic investigation of the

cellulose allomorphs: Cellulose(am), cellulose Iβ(cr), cellulose II(cr), and cellulose III(cr).

J. Chem. Thermodyn., 81, 184–226 (2015).

[51] M. Ioelovich. Heat effects of interaction between cellulose and various polar liquids. SITA,

13, 35-44 (2011).

[52] M. Ioelovich; Study of thermodynamic stability of various allomorphs of cellulose. J. Basic

Appl. Res. Int., 16(2), 96-103 (2016).

[53] M. Ioelovich, V.P. Karlivan. Effect of structural characteristics on reactivity of cellulose.

Wood Chem., 1, 18-25 (1986).

[54] M. Ioelovich. Crystalline structure of alkali-celluloses and its influence on viscose forming

process. Wood Chem., 2, 8-15(1990).

[55] N.I. Klenkova. Structure and Reactivity of Cellulose. Science, 1976.

[56] M. Wada. In situ observation of the crystalline transformation from cellulose IIII to Iβ

Macromolecules, 34, 3271-3275 (2001).

[57] G. Kaptay. The Gibbs equation versus the Kelvin and the Gibbs-Thomson equations to

describe nucleation and equilibrium of nano-materials. J. Nanosci. Nanotech., 12, 1-9

(2011).

[58] M. Ioelovich, R. Luksa. Alteration of crystalline structure of cellulose during thermal

treatment. Wood Chem., 3, 18-20 (1990).

ISOPHASE AND PHASE TRANSITIONS OF CELLULOSE – A SHORT REVIEW

Documents