Journal of Fire Sciences http://jfs.sagepub.

com/

Flame Retardancy of Wool Fabric with Zirconium Oxychloride Optimized by Central Composite Design
M. Forouharshad, M. Montazer, M. Bameni Moghadam and O. Saligheh Journal of Fire Sciences 2010 28: 561 originally published online 14 June 2010 DOI: 10.1177/0734904110373205 The online version of this article can be found at: http://jfs.sagepub.com/content/28/6/561

Published by:
http://www.sagepublications.com

Additional services and information for Journal of Fire Sciences can be found at: Email Alerts: http://jfs.sagepub.com/cgi/alerts Subscriptions: http://jfs.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jfs.sagepub.com/content/28/6/561.refs.html

>> Version of Record - Oct 18, 2010 OnlineFirst Version of Record - Jun 14, 2010 What is This?

Downloaded from jfs.sagepub.com by guest on January 9, 2013

Flame Retardancy of Wool Fabric with Zirconium Oxychloride Optimized by Central Composite Design
M. FOROUHARSHAD*
Textile Department, Islamic Azad University South Branch, Young Researchers Club, Tehran, Iran

M. MONTAZER
Textile Department, Center of Excellence in Textile, Amirkabir University of Technology, Hafez Avenue, Tehran, Iran

M. BAMENI MOGHADAM
Department of Statistics, Allameh-Tabatabai University, Tehran, Iran

O. SALIGHEH
Textile Department, Islamic Azad University Science and Research Branch Tehran, Iran
(Received March 31, 2010)

ABSTRACT: In this study, the effect of zirconium oxychloride (ZrOCl2) with different concentrations of citric and formic acids on the flame-retardant properties of wool is assessed by thermal analysis, mass loss, the limiting oxygen index (LOI) and the vertical flame. Scanning electron micrographs have been prepared to observe and compare the surface morphology of the raw and treated samples. It has been observed that the ZrOCl2 exhausted onto wool and the flame retardancy natural markedly improved. The wool treated with the flameretardant agents show an increase in the temperature of decomposition and residual mass and LOI compared with the untreated wool. The central composite design (CCD) is used for the experimental design with four variables.

*Author to whom correspondence should be addressed. E-mail: Forouharmahdi@gmail.com JOURNAL

OF

FIRE SCIENCES, VOL. 28 November 2010

561

0734-9041/10/06 056112 $10.00/0 DOI: 10.1177/0734904110373205 The Author(s), 2010. Reprints and permissions: http://www.sagepub.co.uk/journalsPermissions.nav

Downloaded from jfs.sagepub.com by guest on January 9, 2013

562

M. FOROUHARSHAD ET AL.

The statistical analysis confirmed the optimum conditions obtained by the experimental results, in which the processing optimized at 94.98C with 9.74% zirconium oxychloride, 9.6% citric acid, and 6.3% formic acid. KEY WORDS: flame retardant, wool, zirconium oxychloride, central composite design.

INTRODUCTION THE FLAMMABLE NATURE of fibrous products is one of the major problems of the present time, fires from textiles causing many deaths and injuries and considerable financial losses. Hazards from flammable fabrics have been recognized for many centuries, and repeated attempts have been made to control them [1]. Wool, a cutin fibrous protein, contains many kinds of cysteine, thiocarbamic acid, and cross-linking polypeptides with a helical structure. The natural flame-resistant properties of wool are connected with its relatively high nitrogen content (16%), high moisture content (1014%) [2], high ignition temperature (5706008C) [3], low heat of combustion (20.5 kJ/g), low flame temperature (6808C), and a relatively high limiting oxygen index (LOI) (2528%) [4]. Wool, when heated alone, pyrolyses by a complex series of reactions, which yield a number of products at increasing temperatures. Initially at 2302408C, rupture of the helical structure occurs and the major ordered part of the wool protein undergoes a solid to liquid phase change [5]. At 2502958C, an endothermic reaction occurs associated with release of sulfur compounds due to the breaking of the cystine disulfide bonds and simultaneous release of hydrogen sulfide. Above 2508C, general pyrolytic decomposition occurs, including char-forming reaction with dehydration and loss of other volatiles. In the presence of air, formation of sulfur dioxide occurs between 2708C and 3208C [5]. Cleavage of the cystine disulfide bond is seen to play a very important role in the thermal degradation and combustion of keratin and it has been suggested that the oxidation of cystine may be the initial exothermic reaction in the burning of wool [5]. With the development of science and technology and the improvement of the standard of life, improvement of the natural flame resistance of wool has provoked peoples interest. In recent years, there have been a number of reports of treatments which enhance wools natural flame-resistant properties [611]. Benisek when working at the International Wool Secretariat Laboratories has observed that mordanting treatments based on zirconium and titanium salts markedly improve wools flame resistance [611]. They used vertical flame test and differential scanning calorimetric (DSC) and

Downloaded from jfs.sagepub.com by guest on January 9, 2013

Flame Retardancy of Wool Fabric

563

thermogravimetry analysis (TGA) to study the flame resistance and thermal behavior. Also, scanning electron microscopy (SEM) and energy dispersive X-ray microanalysis (EDXS) were used to study the morphology of treated wools. Here, in this research, zirconium oxychloride (ZrOCl2) along with two different organic acids (citric and formic acids) were used to produce a FR wool fabric. The FR processing was designed according to response surface methodology (RSM) by adapting the central composite design (CCD) and the processing parameters were optimized through the application of CCD.

EXPERIMENTAL DETAILS Materials The wool fabric with plain woven structure from 48/2 Nm yarns was supplied by Iran Merino. The fabric was scoured with 0.5% nonionic detergent at 50 for 30 min (L : R 40 : 1) and then washed with tap water, and dried at room temperature. The zirconium oxychloride (35%ZrO2) used in this study was supplied by Shanghai Yancui Co, China. Formic acid and citric acid were obtained from Merck, Germany. Preparation of Flame-retardant Bath Formic acid (HCOOH) and citric acid (C6H8O7) were mixed with ZrOCl2 according to the central composite design matrix in Table 2, after which water was added until the solution achieved a 20 : 1 ratio of liquor to wool. Flame-retardant Processing of Wool Formic acid was added to each of the above flame retardants in order to maintain a pH of 3 during the exhaustion procedure. Mordanting of samples was started at 408C for 20 min and the temperature was raised for 30 min to the specified temperature in the CCD matrix (Table 2) and heated for 45 min. After being exhausted, the treated samples were rinsed with tap water and dried at room temperature. The criterion for flame retardation in the present work is that the fabric must pass the vertical strip-test prescribed by the United States Federal Aviation Administration (F.A.A test) [10]. Briefly, this test requires the burning of a vertically held fabric in a draft-free cabinet, in accordance with Federal Test Method Standard 191, Method 5903. A minimum of three specimens must be tested and the results averaged.

Downloaded from jfs.sagepub.com by guest on January 9, 2013

564

M. FOROUHARSHAD ET AL.

A flame 38 mm (1.5 in.) high is applied to the fabric, which is held 19 mm (1.5 in.) above the top edge of the burner. The flame is held in position for 12 s and then removed. For a sample to pass the test, the average burn length must not exceed 20 cm (8 in.), and the average flame time after removal of the flame source must not exceed 15 s. Limiting Oxygen Index The limiting oxygen index (LOI) value is the minimum amount of oxygen in an oxygennitrogen mixture required to support complete combustion of a vertically held sample that burns downward from the top. The higher the LOI value, the more efficient is the flame-retardant treatment. The LOI values were determined in accordance with ASTM D2863-06 by means of a General Model Stanton Redcroft FTA Flammability unit. Thermogravimetry Analysis Trials were performed in a TGA-PL thermogravimetric analyzer (Polymer Laboratories, UK) using a platinum crucible of 70 mL. The following analytical conditions were used: sample mass: 5 mg; initial temperature: 308C; final temperature: 6008C; heating rate: 108C min1; puring gas: nitrogen 50 mL/min. Differential Scanning Calorimetry Trials were performed in a DSC-Maia-200 F3 unit (Netzsch, Germany). Micropunched aluminum pans of 40 mL were used. Condition: sample mass: 13 mg; initial temperature: 208C; final temperature: 6008C; heating rate: 108C/min; puring gas: nitrogen 50 mL/min. Scanning Electron Microscopy (SEM-EDXS) Morphology of the samples was observed by scanning electron microscopy (Philips XL30, The Netherlands). Samples were fixed to SEM holders and coated with a thin layer of gold prior to SEM-EDXS investigation. Condition: acceleration tension: 16 kV; working distance: 9 mm; analysis live time: 100 s. Experimental Design and Analysis The central composite design used for experimental plan with four variables is shown in Table 1. Four variables including ZrOCl2, formic acid and citric acid amount, and temperature were studied. The ranges

Downloaded from jfs.sagepub.com by guest on January 9, 2013

Flame Retardancy of Wool Fabric Table 1. Range of variables.

Variable Temperature (8C) Zirconium oxychloride (ZrOCl2) (%) Citric acid (C6H8O7) (%) Formic acid (HCOOH) (%) Lower limit 77 5.60 6.30 5.65

565

Upper limit 95 10.30 12.80 10.35

Table 2. Central composite design for flame retardant wool.

Run number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 A: Zirconium oxychloride (%) 10.30 10.30 4.00 5.60 10.30 7.95 5.60 11.90 7.95 5.60 7.95 7.95 7.95 7.95 5.60 7.95 7.95 7.95 10.30 7.95 7.95 B: Temperature (8C) 95 77 86 95 95 101 95 86 86 77 86 86 86 86 77 86 71 86 77 86 86 C: Citric acid (%) 12.8 12.8 9.5 12.8 6.3 9.5 6.3 9.5 15.0 12.8 9.5 9.5 9.5 4.1 6.3 9.5 9.5 9.5 6.3 9.5 9.5 D: Formic acid (%) 5.6 10.3 8.0 10.3 5.6 8.0 10.3 8.0 8.0 5.6 8.0 8.0 8.0 8.0 5.6 8.0 8.0 11.9 10.3 4.0 8.0 Char length (cm) 1.2 1.65 1.7 2.5 1 1 1.65 1.35 1.3 1.7 1.75 0.95 0.95 2.1 1.95 0.9 1.3 1.2 1.25 1.6 0.85

of these variables are shown in Table 1. Details of CCD for mordanting of wool with ZrOCl2 are demonstrated in Table 2. Also, the influence of the variable on the results C.L. (char length (cm)) is adjusted using the following second-order polynomial function (Equation (1)): X X X C:L: b0 b i xi bij xi xj C i x2 i i!j i 1,2,3,4

Downloaded from jfs.sagepub.com by guest on January 9, 2013

566

M. FOROUHARSHAD ET AL. Table 3. Regression coefficients and determination coefficient.

C.L. coefficient b0 b1 b2 b3 b4 c1 c2 c3 c4 b12 b13 b14 b23 b24 b34 R2 Char length 13.4 0.68 0.095 0.91 1.45 0.029 3.56E-004 0.0211 0.0212 0.0127 0.0000 0.0116 3.85E-003 0.0110 0.0213 0.724

In this equation, b0 is an independent term according to the mean value of the experimental plan, bi are regression coefficients that explain the influence of the variables in their linear form, bij are regression coefficients of the interaction terms between variables, and ci are the coefficients of quadratic form of variables. Equation regression coefficients bi, bij, ci and the determination coefficient R2 are shown in Table 3.

RESULTS AND DISCUSSION Thermal Behavior Comparisons of TG and DSC curves of pure wool and treated wool (wool treated with 9.74% zirconium oxychloride, 9.6% citric acid, 6.3% formic acid at 958C) are shown in Figures 1 and 2. From DSC and TG curves, the kinetic parameters were obtained and are listed in Tables 4 and 5. Each figure indicates that three processes take place in the wool pyrolysis progress. The first, which ends at around 1608C, is an endothermic process and corresponds with the desorption of water physically bound to fiber and dehydration of wool.

Downloaded from jfs.sagepub.com by guest on January 9, 2013

Flame Retardancy of Wool Fabric

120 Pure wool 100 Residual mass (%) 80 60 40 20 0 0 200 400 Temperature (C) 600 800 Treated wool

567

Figure 1. Thermal behaviors of raw wool and treated wool in nitrogen condition (TGA).
1.5 1 DSC (mW/mg) 0.5 0 0.5 1 0 100 200 300 400 500 600 Temperature (C) Pure wool Treated wool

Figure 2. Thermal behaviors of raw wool and treated wool in nitrogen condition (DSC).

The second important process coincides with the temperature range over which a number of defined pyrolysis reactions take place in wool. The hydrogen-bond peptide helical structure rupture and the ordered regions of the wool undergo a solid to liquid phase change, also cleavage of the disulfide bonds occurs and a number of volatiles are released including hydrogen sulfide and sulfur dioxide [12]. As presented in Figure 1, compared with pure wool, TG curve of treated sample shows a

Downloaded from jfs.sagepub.com by guest on January 9, 2013

568

M. FOROUHARSHAD ET AL. Table 4. The thermal properties of second pyrolysis process of wool with and without flame retardant.
Sample Pure wool Treated woola
a

Temperature range (8C) 190330 212374

Mass loss (%) 38.16 46.05

Sample was obtained with 9.74% ZrOCl2, 9.6% citric acid, and 6.3% formic acid at 958C.

Table 5. The thermal properties of wool with and without flame retardant.
Sample Pure wool Treated woola
a

Temperature range (8C) 391597 403599

Residual mass (%) 30.80 32.02

LOI (%) 25.4 29.3

Sample was obtained with 9.74% ZrOCl2, 9.6% citric acid, and 6.3% formic acid at 958C.

high decomposition temperature and an increased mass loss, which can be seen from Table 4. It is probable that the presence of ZrOCl2 and citric acid, which reacts in the condensed phase, could increase the rate of volatile product formation above 2508C. The third process is an exothermic reaction that the char oxidation reactions dominate [9]. The presence of the flame retardants appears to have resulted in a cross-linked complex, which has achieved a carbonaceous structure, and possibly aromatic char that probably has a high-than-expected resistance to oxidation [12]. The respective mass losses and LOI values of treated wool increases compared with those of untreated wool, which can be seen from Table 5.

Statistical Analysis The analysis of variance (ANOVA) is given in Table 6. According to the ANOVA results, the fitted model is:
C:L: 13:4 0:68 A 0:095 B 0:91 C 1:45 D 0:0127 A B 0:0000 A C 0:0116 A D 3:85E 003 B C 0:0110 B D 0:0213 C D 0:029 A2 3:56E 004 B2 0:0211 C2 0:0212 D2

Downloaded from jfs.sagepub.com by guest on January 9, 2013

Flame Retardancy of Wool Fabric Table 6. ANOVA for response surface quadratic model.
Source Model A ZrOCl2 B Temp C Citric acid D HCOOH AB AC AD BC BD CD A2 B2 C2 D2 Residual Lack of fit Pure error Cor total Sum of squares 2.80 0.061 0.045 1.549E-003 0.080 0.24 0.000 0.014 0.10 0.18 0.21 0.39 0.012 0.74 0.21 1.07 0.50 0.57 3.86 DF 14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 6 2 4 20 Mean square 0.20 0.061 0.045 1.549E-003 0.080 0.24 0.000 0.014 0.10 0.18 0.21 0.39 0.012 0.74 0.21 0.18 0.25 0.14 F-value 1.12 0.34 0.25 8.707E-003 0.45 1.35 0.000 0.077 0.57 1.02 1.19 2.19 0.070 4.19 1.15 1.76

569

p-Value Prob4F 0.4727 0.5787 0.6329 0.9287 0.5274 0.2898 1.0000 0.7910 0.4791 0.3525 0.3176 0.1896 0.8006 0.0867 0.3240 0.2832

In this equation, A, B, C, and D are ZrOCl2 (wt%), temperature (8C), citric acid (o.w.w. %), and formic acid (o.w.w. %), respectively. Figure 3 also shows the response surface of the model. By using design of expert software the optimum design point with desirability of 100% is 94.98C temperature, 9.74%, zirconium oxychloride, 9.6% citric acid, and 6.3% formic acid. Scanning Electron Microscopy (SEM-EDXS) SEM was utilized to study the effect of ZrOCl2 and citric acid on the wool fiber surfaces. The results in Figure 4 show a normal morphological form of the raw wool and scales can be clearly seen (Figure 4(a) and (b)). Images show the decomposition of scales of treated wool. For the mordanted wool, some particles of mordant has been observed on the fiber surface (Figure 4(c) and (d)). The wool sample treated with 9.74 wt% of ZrOCl2 indicated an aggregation of zirconium on the edge of scales and scattering on the fiber surface randomly. Existence of zirconium element and other elements on the surface of the wool treated is investigated by EDXS analysis and is reported in Table 7.

Downloaded from jfs.sagepub.com by guest on January 9, 2013

570
Design-Expert Software C.L 2.5 0.85 X1 = A: ZrOCl2 X2 = B: Temp Actual Factors C: Citric acid = 9.55 D: Formic acid = 8.00 C.L. 1.75 1.51 1.27 1.04 0.8 95.00 90.50 86.00 B: Temp 81.50

M. FOROUHARSHAD ET AL.

10.30 9.13 7.95 6.78 77.00 5.60 A: ZrOCl2

Figure 3. Design of expert plot for treated wool fabric with zirconium oxychloride.

Figure 4. SEMs of (ab) raw wool (cd) wool treated with 9.74% ZrOCl2, 9.6% citric acid, 6.3% formic acid at 958C (L : R 20 : 1).

Downloaded from jfs.sagepub.com by guest on January 9, 2013

Flame Retardancy of Wool Fabric Table 7. Elements analyzed of wool treated (normalized) (wt %).
Sample Treated wool
a

571

Zr
a

S 2.94

Cl 0.27

Au 91.82

Total 100.00

4.97

Sample was obtained with 9.74% ZrOCl2, 9.6% citric acid, and 6.3% formic acid at 958C.

CONCLUSION The treatment of wool with ZrOCl2 and citric acid accelerates the formation of volatiles from the fiber. The thermal behaviors of wool fiber by TGA revealed three major regions of weight loss. During the progress of thermal degradation, ZrOCl2 and citric acid effect on the pyrolysis reaction, giving also higher decomposition temperatures and greater mass losses in the second stage. The results of vertical flame test indicate that the treatment of wool with ZrOCl2 increases the flame retardancy of wool to a great extent. The statistical analysis by DOE indicated that application of 9.74% ZrOCl2, 9.6% citric acid, and 6.3% formic acid at 958C on wool produces an optimum design point with desirability of 100% to give minimal char length (0.84 cm) and maximum LOI (29.3%) values. In SEM images, particles of zirconium on the edge of scales and the change of morphology in comparison with the raw wool are clearly observed.

REFERENCES
1. Popescu, C., Vasile, M., Oprea, C. and Segal, E. (1992). A Thermogravimetric Study of Flame-proofed Wool, Thermochimica Acta, 205: 205211. 2. Benisek, L. (1974). Improvement of the Natural Flame Retardants of Wool Part I: Metal-Complex Applications, Journal of Textile Institute, 65: 102108. 3. Benisek, L. (1974). British patent 1372694. 4. Horrock, A.R. (1986). A Review of Flame-retardant Finishing of Textile, Journal of the Society of Dyers and Colourists, 16: 62101. 5. Horrocks, A.R. (2000). Char Formation in Flame-retarded Wool Fibres. Part 1. Effect of Intumescent on Thermogravimetric Behaviour, Journal of Fire and Materials, 24: 151157. 6. Benisek, L. (1971). Use of Titanium Complexes to Improve the Natural Flame Retardancy of Wool, Journal of the Society of Dyers and Colourists, 87: 277278. 7. Benisek, L. (1972). New Aspects of Flame Protection Using Wool, Textile Manufacturer (Manchester), 99: 3639.

Downloaded from jfs.sagepub.com by guest on January 9, 2013

572

M. FOROUHARSHAD ET AL.

8. Benisek, L. (1972). New Aspects of Flame Protection Using Wool Versatile, Simple, Inexpensive, Int. Dyer Text. Printer, Bleacher Finish, 147: 414419. 9. Tian, C.M., Li, Z., Guo, H.Z. and Xu, J.Z. (2003). Study on the Thermal Degradation of Flame Retardant Wools, Journal of Fire Sciences, 21: 155162. 10. Tian, C.M., Zhang, H.Y., Xu, J.Z., Gua, H.Z. and Shi, J.R. (1998). Studies on the Flame Retardation and Thermal Degradation of Wool, Journal of the Textile Institute, Part 13, 89: 591594. 11. Federal Register, 37 (37), February 24, 1972, p. 3972. 12. Davies, P.J., Horrocks, A.R. and Miraftab, M. (2000). Scanning Electron Microscopic Studies of Wool/Intumescent Char Formation, Polymer International, 49: 11251132.

Downloaded from jfs.sagepub.com by guest on January 9, 2013