Refining of soda-AQ, kraft-AQ, and ethanol pulps from orange tree wood

Abstract

The pulp yield of orange tree wood was tested under various conditions including processing with soda-anthraquinone (soda-AQ), kraft-anthraquinone (kraft-AQ), or ethanol under different temperature, time, reagent concentration, and PFI laboratory beater beating regimes. Beating grade and stretch properties were studied, with a view to identifying the optimum operating conditions. Polynomial equations were derived that generally reproduced the dependent variables, with errors in most cases much less than 20%. Kraft-AQ pulping was the most efficient. The values of the tensile, burst, and tear indices obtained with kraft-AQ (78.04 Nm/g, 4.84 kN/g, and 2.97 mNm²/g, respectively), were in most cases higher than those found for soda-AQ and ethanol pulps. Using lower values of operational conditions than those required to maximize the studied paper properties (170 °C, 65 min, 13% active alkali, and 2700 number of PFI beating revolutions), it was possible to provide a more energy- and chemically-efficient process for industrial facilities.

Full Article

Refining of Soda-AQ, Kraft-AQ, and Ethanol Pulps from Orange Tree Wood

Zoilo González, Alejandro Rodríguez,* Fátima Vargas, and Luis Jiménez

The pulp yield of orange tree wood was tested under various conditions including processing with soda-anthraquinone (soda-AQ), kraft-anthraquinone (kraft-AQ), or ethanol under different temperature, time, reagent concentration, and PFI laboratory beater beating regimes. Beating grade and stretch properties were studied, with a view to identifying the optimum operating conditions. Polynomial equations were derived that generally reproduced the dependent variables, with errors in most cases much less than 20%. Kraft-AQ pulping was the most efficient. The values of the tensile, burst, and tear indices obtained with kraft-AQ (78.04 Nm/g, 4.84 kN/g, and 2.97 mNm²/g, respectively), were in most cases higher than those found for soda-AQ and ethanol pulps. Using lower values of operational conditions than those required to maximize the studied paper properties (170 °C, 65 min, 13% active alkali, and 2700 number of PFI beating revolutions), it was possible to provide a more energy- and chemically-efficient process for industrial facilities.

Keywords: Orange tree wood; Soda-anthraquinone; Kraft-anthraquinone; Ethanol; Pulp refining

Contact information: Chemical Engineering Department, University of Córdoba, 14071, Córdoba, Spain; * Corresponding author: a.rodriguez@uco.es

INTRODUCTION

Since the 1970s, pulp production from non-woody plants using non-conventional raw materials has increased from approximately 7% to almost 12% of the total pulp produced, growing at a rate of 2 to 3 times greater than wood pulps (Atchinson 1996; Simula 2002; Alaejos et al. 2004; González et al. 2011).

The use of agricultural and agro-industry residues and alternatives to food crops seems to be a good alternative to raw wood material, which can lead to excellent paper products with special properties and can serve as the sole source of raw materials in some geographical areas (Jiménez 2005).

Orange tree prunings could provide a new source of non-wood raw material. Spanish production of orange tree prunings from felling operations is more than 5 million tons per year (Rodríguez et al. 2010). Orange tree prunings and crop residues in general must be removed to control pollution, fire, pests, interference with soil cultivation, and occupation of large areas. It is advantageous to try to exploit the different fractions of waste as a method to reduce disposal costs.

Many non-wood raw materials such as orange tree prunings contain fractions unsuitable for the production of pulp, including the leaves, bark, pith, and young stems, which contain relatively low cellulose content. However, these fractions, which can be called residue, could be used as fuel for producing heat for heating (Arvelakis and Koukios 2002; Ozturk and Bascetinlik 2006; Overend and Wright 2005; González et al. 2012).

The main fraction of orange tree prunings, such as branches and stems with diameters larger than 1 cm, could be used to produce cellulosic pulps for paper. A single study related to pulping of orange tree prunings was found in the literature (González et al. 2011). In this work, a chemical characterization of orange tree prunings (α-cellulose, holocellulose, and lignin contents) was conducted. The group also studied the variables associated with soda-AQ pulping on pulp and papersheet characteristics. The conclusion of this work was that this raw material has potential for cellulosic pulp production.

The aim of the present work was to study and compare three pulping processes, soda-AQ, kraft-AQ, and ethanol, applied to the main fraction of orange tree prunings, as well as the refining of the pulps obtained. The influence of the operational variables, temperature, time, reagent concentrations, and number of PFI beating revolutions, on the characteristics of the papersheets obtained was studied to determine the optimal operating conditions.

EXPERIMENTAL

Raw Material

This work used the main fraction of orange tree (Citrus sinensis) prunings, which consisted of the wood from branches and stems with diameters larger than 1 cm. The orange wood was characterized according to TAPPI standards, namely, T9m 54, T203 os-61, T222, T211, and T204 for holocellulose, α-cellulose, lignin, ash, and ethanol-benzene extractives, respectively.

Pulping and Pulp and Paper Characterization

Pulps were obtained using a 15-L batch cylindrical reactor. The raw material was cooked in the reactor using soda-AQ, Kraft-AQ, or ethanol. Next, the cooked materials were fiberized in a wet disintegrator at 1200 rpm for 30 min, and the screenings were separated by sieving through a screen of 0.14-mm mesh size. The pulps were beaten on a PFI refiner from Metrotec with precise control of the number of beating revolutions used. The drainability or beating grade (Shopper-Riegler index) of the pulps was determined according to TAPPI T220 sp-96. Pulp yield was determined by weighing, after removing the uncooked materials.

Paper sheets were prepared with an ENJO-F-39.71 sheet machine by Metrotec according to the TAPPI T205 ps-95 standard. The tensile index, burst index, and tear index of paper sheets were determined according to TAPPI standards T494 om-96, T403 om-97, and T414 om-98, respectively.

Experimental Design

The following procedure was carried out to quantify the effects of the operational variables (3 variables of pulping process and 1 of the refining process): a 2ⁿ factorial design was used for the three pulping process variables, consisting of a central experiment (in the centre of a cube) and 14 additional points (additional experiments lying at the cube vertices and side centers) (Montgomery, 1991); then, the pulps of each of those 15 experiments were subjected to 3 refining experiments, resulting in a total of 60 experiments (with 15 experiments without refining).

RESULTS AND DISCUSSION

The chemical characterization of orange tree wood, as well as other similar non-wood materials and wood of pine and eucalyptus is presented in Table 1.

Table 1. Chemical Characterization of Wood an Non-wood Materials

The orange tree wood holocellulose content is higher than the other materials considered with the exception of eucalyptus. The α-cellulose content is higher than the content of olive tree and vine shoots, but less than other materials. The lignin content is similar to the materials considered with the exception of the pine. The extractives are high compared with the cotton stalks, pine and eucalyptus, but lower than the other materials. The ashes are high, as in the case of vine shoots, in comparison with other materials (Table 1).

Operating conditions in the three experimental designs applied to three pulping processes of orange tree wood (soda-AQ, kraft-AQ and ethanol) are presented in Table 2.

Table 2. Experimental Conditions used in the Soda-AQ, Kraft-AQ, and Ethanol Pulping Applied to Orange Tree Wood

Table 3 shows the normalized values of the operational variables for the 15 experiments of each experimental design, as well as the average values of three experimental values of the pulp yield.

Table 3. Normalized Values of the Pulping Operational Variables and Experimental Average Values (of 3 experiments) of the Pulp Yield for Different Pulps from Orange Tree Wood

Fifteen pulps were obtained from each of the experimental designs and were refined in three different tests with a number of PFI beating revolutions of 1000, 2000, and 3000. Table 4 shows the values of the operational variables to obtain different refined pulps, as well as the average values of six trials of the beating grade of the pulps and of the tensile, burst, and tear indices of the paper sheets.

Table 5 shows the equations obtained to fit the experimental data of the Table 4 with a polynomial model.

For pulp yield, beating grade and tear index (in the three pulping processes) and for tensile index (in the soda-AQ and kraft-AQ pulping process), the values estimated using the equations in Table 3 reproduced the experimental results with errors smaller than 20% in all cases, being in most cases much less than 20%. Figure 1, for the beating grade in the soda-AQ pulping, confirms what is stated above. For the other dependent variables similar graphics were obtained.

For the burst index (in the three pulping processes) and for the tensile index in the ethanol pulping, polynomial models were not suitable because the values estimated from these variables deviated greatly from the experimental values

Using non-linear programming as implemented in the More and Toraldo method (1989), it was possible to identify the values of the operational variables providing the greatest values of the dependent variables for the pulp and paper sheets (Table 6). It can be seen that for each pulping process, the operating conditions to get the maximum value from a given dependent variable are different. Thus, for example, for soda-AQ pulping, if the paper sheets require a high tensile index (a high breaking length), the pulp should be obtained with medium temperature and a medium-high soda concentration, in addition to a high number of PFI beating revolutions in the pulp refining.

Table 7 shows the highest values of the variations of the dependent variables (obtained from the equations in Table 5, by varying each of the operational variables and keeping the values of the remaining operational variables with their optimal values). Table 7 also shows the maximum deviation of the dependent variables (in %) with respect to their optimal values due to the maximum variations calculated previously.

Fig. 1. Values predicted versus experimental for the beating grade (°SR)

From Table 7, it can be concluded that, in the soda-AQ and kraft-AQ pulping, the most influential operational variable on the strength properties of paper was the number of PFI beating revolutions. By contrast, the pulping time affected stretch the least; in the case of ethanol pulping, the most influential variable was also the number of PFI beating revolutions, but the least influential factor was the ethanol concentration.

On the other hand, comparing the results in Table 4, for soda-AQ and kraft-AQ pulping, the burst index values corresponding to the different pairs of experiments that are differentiated only by the value of one of the operational variables (keeping the other three operational variables equal in each of the two compared experiments) verify that the most influential variable on the value of the burst index was the number of PFI beating revolutions and the least influential was the pulping time; for ethanol pulping, the least influential variable on the burst index was ethanol concentration. For ethanol pulping (see Table 4), the most influential variable for the tensile index value was PFI beating revolutions, while the ethanol concentration and the pulping time were less important.

Table 4. Normalized Values of the Pulping and Refining Variables and Experimental Average Values (of 6 experiments) of the Paper Properties for Different Pulps from Orange Tree Wood

Table 5. Polynomial Equations for the Dependent Variables in the Pulping and Refining from Orange Tree Wood Using Different Pulping Processes

Table 6. Operational Variable Values in Orange Tree Wood Pulping and Refining to Obtain Optimal Values of Dependent Variables

Table 7. Maximum Variation Values of the Dependent Variables Varying one of the Operational Variables in Orange Tree Wood Pulping and Refining

Table 8 shows the results of the simulation of the pulping and refining through the equations of Table 5. By use of these operating conditions acceptable values for the strength properties of the pulps were obtained, and at the same time they deviated very little from their maximum values (which are shown in Table 6). Moreover, there was only a small fall-off in the values of yield and beating grade. This mode of operation discards both mild and severe operating conditions. Mild operating conditions give rise to strength properties of the pulps that are too low, while severe operation conditions give rise to low pulp yield and excessive consumption of energy (operating at high temperature and number of PFI beating revolutions) and reagents (operating with high concentration), as well as a high capital assets for installation (operating for a high pulping time).

Table 8. Orange Tree Wood Pulping and Refining Simulation

The polynomial models obtained for the different dependent variables were similar to those previously reported for paper sheets from some types of agricultural and agro-industries residues: acetone, ethanol, and ethanol-acetone pulp from wheat straw (Jiménez et al. 2001, 2002 and 2004); kraft pulp from olive wood pruning (López et al. 2000, Díaz et al. 2005); and soda-anthraquinone pulp from empty fruit bunches of oil palm (Jiménez et al. 2009).

Table 9 shows the optimal results obtained in this work for orange tree wood pulps, as well as those of other studies for pulps beaten from wheat straw, olive wood, and empty fruit bunches (EFB) of oil palm.

Table 9. Comparison of Various Pulps Beaten from Non-Wood Raw Material

The data in Table 9 suggests that the values of the yield and beating grade of the pulps of orange tree pruning are intermediate to those of the non-wood materials considered. The values of the tensile and burst index notes are similar to those of the comparative materials, while the tear index values are lower. Finally, it can be concluded that orange tree wood produced beaten pulps with strength properties that can compete with and even surpass pulps of three singular raw materials pulped with organosolv, soda, and kraft processes: wheat straw, which is widely known and used in the world, olive wood, which contains a ligneous structure intermediate between hardwood and softwood, and EFB, which is a very abundant and very localized agrifood industry residue.

CONCLUSIONS

1. Orange tree wood has contents of holocellulose (73.2%), α-cellulose (48.0%) lignin (20.0%), extractives (3.6%), and ash (3.4%) of the same order as other non-wood and wood materials used for the production of pulp and paper, so it can be regarded as an alternative raw material.
2. For soda-AQ, kraft-AQ, and ethanol pulping of orange tree wood, polynomial equations were derived that reproduced yield, drainability, and strength properties with errors in most cases much less than 20%.
3. Kraft-AQ pulping can be regarded as the most favorable process because, when operated at 170 °C, 65 min, 13% alkali, and 2700 number PFI beating revolutions, values for the tensile index, burst index, and tear index of 73.98 Nm/g, 3.84 kN/g, and 2.97 mNm²/g, respectively, were obtained, and these values are higher than those usually found for soda-AQ and ethanol pulping.

ACKNOWLEDGMENTS

The authors are grateful to Ecopapel, S. L. (Écija, Seville, Spain) for their support and to Spain’s DGICyT and Junta of Andalucía for funding this research within the framework of the Projects CTQ-2010-19844-C02-01 and TEP-6261.

REFERENCES CITED

Alaejos, J., López, F., and Eugenio, M. E. (2004). “Using of non-wood raw materials for cellulose pulp. Review,” Afinidad 61(513), 400-410.

Arvelakis, S., and Koukios, E. G. (2002). “Physicochemical upgrading of agroresidues as feedstocks for energy production via thermochemical conversion methods,” Biomass Bionerg. 22(5), 331-348.

Atchinson, J. E. (1996). “Twenty-five years of global progress in non-wood plant fiber repumping,” Tappi J. 79(10), 87-97.

Díaz, M. J., Eugenio, M. E., López, F., and Alaejos, J. (2005). “Paper from olive tree residues,” Ind. Crop. Prod. 21(2), 211-221.

González, Z., Feria, M. J., Vargas, F., and Rodríguez, A. (2012). “Comparison of the heating values of various types of fuel from non-wood raw materials,” American J. Environ. Engineer.2(4), 91-96

González, Z., Rosal, A., Requejo, A., and Rodríguez, A. (2011). “Production of pulp and energy using orange tree pruning,” Bioresour. Technol. 102 (19), 9330-9334

Jiménez, L. (2005). Pastas Celulósicas de Materias Primas Alternativas a las Convencionales. Editorial Gráficas Sol, Sevilla.

Jiménez, L., García, J. C., Pérez, I., Ariza, J., and López, F. (2001). “Acetone pulping of wheat straw. Influence of the cooking and beating conditions on the resulting paper sheets,” Ind. Eng. Chem. Res. 40(26), 6201-6206.

Jiménez. L., Pérez, I., García, J. C., López, F., and Ariza, J. (2004). “The influence of the ethanol pulping of wheat straw and of the beating of the pulp on the resulting paper sheets,” Wood Sci. Technol. 38(2) 127-137.

Jiménez, L., Pérez, I., López, F., Ariza, J., and Rodríguez, A. (2002). “Ethanol-acetone pulping of wheat straw. Influence of the cooking and beating of the pulps on the properties of the resulting paper sheets,” Bioresour. Technol. 83(2) 139-143.

Jiménez, L., Serrano, L., Rodríguez, A., and Sánchez, R. (2009). “Soda-anthraquinone pulping of palm oil empty fruit bunches and beating of the resulting pulp,” Bioresour. Technol. 100 (3), 1262-1267.

López, F., Ariza, J., Pérez, I. and Jiménez, L. (2000). “Influence of the operating conditions on the properties of paper sheets obtained by kraft pulping of olive tree wood,” Bioresour. Technol. 72(2) 147-151.

Montgomery, D. C. (1991). Design and Analysis of Experiments. Grupo Editorial Iberoamericana, Mexico.

More, A., and Toraldo, A. (1989). “Algorithms for bound constrained quadratic programming problems,” Numerische. Mathematik. 55 (4), 377-400.

Ozturk, H. H., and Bascetinlik, A. (2006). “Energy, exploitation of agricultural biomass potential in Turkey,” Energy Exploration and Exploitation 24(4), 313-330.

Overend, R. P., and Wright, L. L. (2005). “Biomass energy,” in: CRC Handbook of Mechanical Engineering, Second Edition, CRC Press, Boca Raton, FL, pp. 28-44.

Rodríguez, A, Rosal, A., and Jiménez, L. (2010). “Biorefinery of agricultural residues by fractionation of their components through hydrothermal and organosolv processes,” Afinidad67(545), 14-19.

Simula, M. (2002). “Fibre supply: The achilles heel of the global pulp and paper industry,” Paperi Ja Puu-Paper and Timber 84(1), 16-19.

Article submitted: June 21, 2013; Peer review completed: July 31, 2013; Revised version received: September 3, 2013; Accepted: September 15, 2013; Published: September 23, 2013.