[Home] [APPW 2004] [Journal papers]

 

EFFECT OF RELATIVE HUMIDITY AND UNREACTED AKD ON AKD SIZING

Authors

Hak Lae Lee1 and Philip Luner2

Companies

1 Department of Forest Sciences, College of Agriculture and Life Sciences, Seoul National University, Seoul, 151-742, Rep. of Korea
2 ESPRI, SUNY, ES&F, Syracuse, NY 13210, USA

Keywords

AKD, paper sizing, relative humidity, sizing efficiency

 

 

 

 

ABSTRACT

The role of relative humidity and unreacted AKD molecules on the sizing development of AKD (alkylketene dimer) was investigated using carbon-14 labeled AKD. A solvent sizing method was used to control the amount retained on paper and to achieve even distribution of AKD.

The reaction ratio and sizing degree increased with relative humidity up to 80% due to the favorable exposure of AKD lactone rings for reaction. Higher relative humidity had an adverse effect most likely due to the increased separation distance between AKD and the hydroxyl groups of cellulose.

The contribution of unreacted AKD on sizing efficiency relative to reacted AKD was found to depend on the amount of reacted AKD present in the paper. The maximum sizing efficiency of unreacted AKD was observed when the sheet contained 0.025% of reacted AKD.

Introduction

Alkylketene dimer (AKD) has been widely used in the paper industry as a sizing agent since it was developed in 1950's. AKD is commonly described as a reactive size or alkaline size because it forms covalent bonds with cellulose and neutral or alkaline pH values are employed in the papermaking system. Many papers have been published on the mechanism of sizing development and papermaking variables including retention and drying affecting AKD sizing (Lindström and Söderberg 1986; Roberts and Garner 1985).

There are still many unanswered questions and controversy on AKD sizing. For instance, the formation of covalent bond between the hydroxyl groups on cellulose and AKD led to much conjecture and debates. Bottorff (1994) investigated the reactions of AKD occurring in alkaline paper and papermaking systems using solid-state 13C-NMR and provided direct evidence supporting the formation of AKD-cellulose ß-keto esters on AKD sized paper. On the other hand, Isogai (1999) reported that nearly no ß-keto esters formed in the AKD-sized paper using solid-state 13C-NMR.

Discrepancy on the sizing mechanism of AKD led to different views on the role of unreacted or extractable AKD on sizing development. Different results have been reported on the efficiency of extractable AKD on sizing development. Lindström and Söderberg (1986) showed that for most of the pulps the sizing efficiency of the reacted AKD is between 2-3 times the efficiency of the unreacted AKD. On the other hand, Marton (1990) reported that Hercules Size Tester (HST) data for AKD sized sheets did not change significantly after chloroform extraction that removed 45-60% of the retained AKD material. In these paper samples the content of reacted AKD ranged from 0.05-0.09%. Chloroform extract of AKD sized paper contained only ketone, and alkaline hydrolysis of the extracted sheets completely destroyed hydrophobicity, suggesting that the high HST value in the extracted sheets was due to hydrolysable ester bonds (Marton 1990). On the other hand, Isogai (1999) reported that ketone molecules, which are formed from AKD by hydrolysis, are likely to contribute to the appearance of sizing features.

It is required for AKD molecules to approach in close proximity to the hydroxyl groups of cellulose if any reaction would occur. Water molecules absorbed on cellulose surface or alkyl chains of AKD, however, may prevent the functional groups from making chemical contacts.

The purpose of this study was to expand our understanding of the role of moisture levels of cellulose fibers on AKD reaction and to examine the effect of the unreacted AKD on sizing development.

It is well known that such steps as retention, distribution, orientation, and immobilization of sizing agents are required to produce good resistance to wetting and penetration of liquids on internally sized paper. Use of frequently employed AKD emulsion in studying sizing has some disadvantages. First, it is very difficult to make sheets of paper containing a given amount of AKD with an emulsion since the retention of AKD varies depending on diverse papermaking variables. Second, as a droplet of an AKD emulsion contains many AKD molecules, it is doubtful whether a good molecular distribution of AKD can be obtained even at high temperature drying (Ganier and Yu 1999).

To minimize the problems related to AKD emulsion sizing, a solution application method can be used. In addition, hydrolysis leading to by-products as well as entrapping of AKD emulsion particles between fibers (Lindström and Söderberg 1986) can be eliminated by employing solvent sizing.

Materials and methods

Carbon-14 labeled tetradecyl ketene dimer was synthesized with 1-14C hexadecanoic acid, unlabeled hexadecanoic acid, and thionyl chloride according to the methods described in the literature (Roberts and Garner 1985) and recrystallized. Commercial AKD (Aquapel 364, Hercules Inc.) made from a mixture of palmitic and stearic acid (55:45) was recrystallized three times from acetone.

Toluene was used to prepare the radioactive AKD solution and tetrahydrofuran (THF) was used to extract unreacted AKD from the substrate. A 0.1% stock solution of AKD was prepared by dissolving the labeled and normal AKD (50:50) in toluene. A given amount of this AKD solution was applied to Whatman No. 1 filter paper with an Eppendorf constant volume pipetter. Half of the solution was applied to each side of a paper strip to obtain even distribution. Sized paper strips were cured at room temperature or in an oven after air drying . Oven drying was used since migration of the unreacted AKD was noticed from the sized sheet to the blotters or plates touching the AKD sized paper in press drying. Soxhlet extraction was carried out with THF for 12 hours to extract unreacted AKD from the paper, and the amount of unextractable AKD was determined by a Packard TriCarb liquid scintillation counter.

Constant relative humidities were obtained using various saturated salt solutions. A modified water immersion test was used to measure the water absorption of paper. A paper sample was immersed in distilled water for 30 seconds and excess water on the paper surface was removed with blotters. The percentage of water absorption was calculated from the ratio of the mass of water absorbed to the mass of the conditioned test piece.

Results and discussion

Use of a solvent sizing method was inspired since it would allow easy and accurate control of the amount of AKD applied on paper substrate and give even distribution of AKD on its surface. Effects of curing time and moisture on DPM (disintegration per minute) measured by scintillation counting are shown in Table 1. Application of AKD either to Whatman filter papers or directly to counting vials did not cause any variation in DPM. Table 1 also shows that heating did not affect the DPM indicating not only hydrolysis but also thermal degradation of the AKD is negligible under these experimental conditions. AKD molecules appeared not to penetrate into the fiber-fiber bonding area since the DPM's of the AKD sized filter papers peeled into 7 plies or disintegrated into individual fibers were the same as that for control sample. These results showed that with the solvent sizing method papers could be sized with an exact amount of AKD, and undesirable side effects such as entrapping, hydrolysis and uneven distribution could be eliminated.

Table 1. Effects of curing time and moisture on solvent sizing of AKD

Oven drying time   at 100°C (min)

AKD applied to

DPM

0

counting vial

1927

0

paper*

1919

30

counting vial

1905

30

paper

1907

30

paper soaked in water

1918

60

paper

1899

60

counting vial

1915

 

 

average = 1913

* Whatman No. 1 filter paper  

Fig 1 shows the amount of reacted AKD varies with the relative humidity. Maximum AKD reaction was obtained at 80% of relative humidity. Results of the percentage of water absorption for the extracted AKD sized papers showed that an increase in the AKD reaction resulted in a decrease in water absorption (Fig 2). The highest AKD reaction percentage and the lowest water absorption were obtained at 80% relative humidity.

fIGURE 1

Fig 1. Effect of relative humidity and curing time on the reaction percentage of AKD. The amount of AKD applied was 0.01%.

fIGURE 2

Fig 2. Effect of relative humidty on water absorption of AKD sized paper after THF extraction. The amount of AKD applied was 0.01%.

An increase of AKD reaction with relative humidity at low range of relative humidity could be attributed to the increase of surface area accompanied by the well-known swelling of cellulose fiber and the opening up of the collapsed fibril structure of dried fibers (Norberg 1970). However, for water-swollon cotton filter paper, an overall decrease in the external surface area with increasing relative humidity was observed which was attributed to the return of the microfibrils to their equilibrium collapsed state in the presence of sorbed water and capillary condensation (Dorris and Gray 1981). Thus, the increase in the amount of reacted AKD with cellulose as the relative humidity increases to 80% cannot be attributed to the change of surface area of the cotton paper.

It is known that the mobility of the intrinsic side chains plays a significant role in determining the surface properties of reverse phase chromatography supports (Gilpin 1984). According to this concept at a low humidity environment the alkyl chains of AKD would be in a stretched conformation and provide steric hindrance for AKD molecule to react with hydroxyl groups of cellulose (Fig 3a). On the other hand, the long AKD alkyl chains are expected to be in a collapsed conformation in water or in a high humidity environment due to hydrophobic interactions. In addition, under these conditions the hydrophilic lactone rings of AKD molecules are expected to be oriented towards a hydrophilic surface, which would enhance the AKD reactivity to cellulose (Fig 3b). The sudden drop in the AKD reactivity between 80% and 100% relative humidity may be explained by the increased separation distance between hydroxyl groups of the cellulose and the AKD molecules accompanied by multilayer adsorption of water molecules. For cotton cellulose, Stamm (1964) has shown that 5.9-6.9 molecular layers of water are adsorbed at saturation. This multilayer adsorption of water at high relative humidity reduces the number of collisions between AKD molecules and the hydroxyl groups on cellulose, and consequently the amount of reacted AKD (Fig 3c). This suggests that relative humidity has a catalytic effect on AKD sizing by changing the conformation of AKD alkyl chains up to 80% relative humidity.

Figures 3a and 3b

Figure 3c

Fig 3. Effect of relative humidity on the conformation of AKD alkyl chains and AKD reactivity at (a) low relative humidity, (b) high relative humidity, and (c) extremely high relative humidity.

Fig 4 shows the water absorption for unextracted AKD sized papers. The water absorption values for the unextracted papers are less than those of the extracted papers suggesting some contribution of the unreacted AKD on sizing. The decrease in water absorption above 80% relative humidity for the unextracted AKD sized paper was due to the presence of adsorbed water in the paper which increased the moisture content and in turn reduced the water uptake in the immersion test.

Figure 4

Fig 4. Effect of relative humidty on water absorption of AKD sized paper before THF extraction. The amount of AKD applied was 0.01%.

To understand the effect of the unreacted AKD on sizing development a series of experiments were performed using solvent sized Whatman filter papers. Since the AKD reaction rate with cellulose as well as the hydrolysis rate changes with relative humidity, a standard temperature of 23°C and 50% relative humidity were selected as the curing conditions.

Fig 4 shows the changes in water absorption of the unextracted AKD sized papers as a function of AKD addition and curing time. Water absorption of 25% was appeared to be practical minimum for AKD sized filter papers. Evidently, addition of AKD above a critical level did not reduce the water absorption.

Water absorption values as a function of curing time for the extracted AKD sized paper samples are shown in Fig 5. In general, water absorption increased after extraction except for the papers treated with 0.089% AKD and cured for more than 100 hours. These papers maintained the minimum water absorption level of 25% after THF extraction. The most pronounced increase in water absorption after THF extraction was observed for the sample sized with 0.068% AKD.

Figure 5

Fig 5. Water absorption for unextracted paper as a function of the amount of AKD applied and curing time.

In Fig 6 the change of water absorption was depicted as a function of the amount of reacted AKD for the extracted papers shown in Fig 5. In this figure, 200 DPM is equal to 0.01% of AKD. Fig 6 shows that a threshold amount of reacted AKD, which is around 0.01%, is required to obtain any sizing effect. Fig 6 also indicates that the water absorption varies linearly with the amount of reacted AKD in the range between 0.01 and 0.04% of AKD. After 0.04% no change in water absorption was observed. The slope of this linear portion may be regarded as the sizing efficiency of the reacted AKD.

Figure 6

Fig 6. Water absorption for AKD sized papers after THF extraction as a function of the amount of AKD applied and curing time.

Since the sizing efficiency of unreacted AKD depends on the amount of reacted AKD as shown in Fig 4 and Fig 5, to obtain the sizing efficiency of unreacted AKD it is required to prepare papers containing a certain amount of reacted AKD with different levels of unreacted AKD. Consecutive steps consisted of the application of AKD solution for solvent sizing followed by curing and extraction were conducted onto the same paper samples to investigate the role of unreacted AKD on sizing development. After each step of the application of AKD solution the paper strip was either air dried or cured for 30 min in an oven at 105°C. The amounts of total AKD (reacted and unreacted) and reacted AKD after THF extraction were measured by liquid scintillation counting. Fig 7 shows the water absorption values as a function of the AKD percentage for these samples. The number in front of A (application) or E (extraction) in Fig 7 indicates the number of AKD solution applications. For instance, 3E means that three cycles of AKD solution application, curing, and THF extraction have been made for this paper strip, hence it contains only reacted AKD. Notations with a number and A in the rear indicate that these papers contain reacted as well as unreacted AKD. The amount of the unreacted AKD for these papers is equal to the amount of the AKD applied in the final application step since AKD does not react with cellulose during air drying. After oven curing, the total AKD amount remains constant while the amount of reacted AKD increases. The amount of unreacted AKD after curing can be determined by subtracting the DPM after THF extraction form the DPM before extraction. Thus three paper samples could be prepared containing the same amount of reacted AKD and two different amounts of unreacted AKD. For instance, three samples of 2E, 3A, and 2AC (2A after curing) contained equal amount of reacted AKD. They contained, however, different amounts of unreacted AKD. The thick solid line in Fig 7 indicates the average water absorption of the extracted papers that contain reacted AKD only. The shape and inflexion points coincide with those in Fig 6. The sizing efficiency of reacted AKD was obtained from the slope of the mid-section of the thick solid line from 1E to 5E. The water absorption of papers containing reacted and unreacted AKD are shown by unfilled squares. Three water absorption values for papers containing the same amount of reacted AKD gave almost straight lines as shown in Fig 7. Thus, from the slope of these lines the sizing efficiency of unreacted AKD was determined. Since the slopes approached zero at extremely low and high reaction level of AKD, corresponding to the threshold and maximum sizing level, respectively, they were not shown in Fig 7.

Figure 7

Fig 7. Water absorption of extracted AKD sized papers.

The sizing efficiency of unreacted AKD relative to reacted AKD was calculated by dividing the slope for total AKD by the slope for reacted AKD, and shown in Fig 8. This figure shows clearly that the efficiency of the unreacted AKD depends upon the amount of reacted AKD on paper. A maximum relative sizing efficiency of unreacted AKD was obtained at 0.025% reacted AKD. At lower or higher sizing levels of reacted AKD, the sizing efficiency of unreacted AKD decreased.

Figure 8

Fig 8. Effect of unreacted AKD on water absorption. Thick solid line indicates the change of water absorption with the amount of reacted AKD only. Unfilled squares are for the paper with reacted and unreacted AKD. 8 µg or 13 µg (underlines) of AKD was applied in each application step. Curing: air dry. Oven heated 20 min at 105°C.

Figure 9

Fig 9. Relative sizing efficiency of unextracted AKD against the amount of reacted AKD.

The dependence of the sizing efficiency of unreacted AKD on the reacted AKD amount may be explained in terms of the molecular arrangement of AKD. The unreacted AKD molecules may be organized around reacted AKD by hydrophobic attraction. The presence of a critical amount of reacted AKD appears to be necessary for this organization. If too little is present, the distance between reacted AKD molecules is so large that the mutual interactions between AKD alkyl chains cannot occur. Unreacted AKD molecules therefore will be overturned or coiled up rather easily when they contact water molecules. The water absorption values do not decrease significantly beyond 0.04% of the reacted AKD. In this case unreacted AKD did not contribute to a decrease in water absorption

Conclusions

The role of relative humidity and unreacted AKD molecules on the sizing development of AKD (alkylketene dimer) was investigated using carbon-14 labeled AKD. A solvent sizing method was used to control the amount retained on paper and to achieve even distribution of AKD.

The reaction ratio increased with relative humidity up to 80%. This is due to the favorable exposure of AKD lactone rings for reaction. Higher relative humidity had an adverse effect due to the increased separation distance between AKD and the hydroxyl groups of cellulose. Contribution of the unreacted AKD on water absorption and its sizing efficiency relative to reacted AKD were explored. The sizing efficiency of the unreacted AKD was found to depend on the amount of reacted AKD present on paper. Maximum efficiency of the unreacted AKD was 0.7 at 0.025% of reacted AKD.

Acknowledgements

Financial support for this sudy from the Government of Korea, BK 21 project, ESPRA and IBM are deeply appreciated.

References

Bottorff, K.J. (1994): AKD sizing mechanism: a more definitive description. Tappi J. 77(4): 105.

Dorris, G.M. and Gray, D.G. (1981): Effect of relative humidity on the external area of paper. J. Chem. Soc. Trans. Faraday Soc., I, 77, 713.

Isogai, A. (1999): Mechanism of paper sizing by alkylketene dimers. J. Pulp Paper Sci. 25(7): 251.

Garnier, G., and Yu, L. (1999): Wetting mechanism of a starch-stabilized alkylketene dimmer emulsion: A study by atomic force microscopy, J. Pulp Paper Sci. 25(7): 235.

Gilpin, R.K. (1984): The bonded phase: structure and dynamics, J. Chromatographic Sci. 22, 371.

Lindström, T. and Söderberg, G. (1986): On the mechanism of sizing with alkylketene dimmers, Part 1: Studies on the amount of alkylketene dimmers required for sizing different pulps, Nordic Pulp Paper Res. J. 1(1):26.

Marton J. (1990): Practical aspects of alkaline sizing: On kinetics of alkyl ketene dimmer reaction: hydrolysis of alkyl ketene dimer. Tappi J. 73(11):39.

Norberg, P.H. (1970): Electron microscopy observations of the effects of drying on cellulose fibers, Svensk Papperstid. 73(7).:208.

Roberts, J.C. and Garner, D.N. (1985): The mechanism of alkyl ketene dimmer sizing of paper, Part 1, Tappi J. 68(4):118.

Stamm, A.J. (1964): Wood and Cellulose Science, New York, Ronald Press, 192.

BACK TO TOP