Cellulose chars

ESR Studies. Jackson and Wynne-Jones (9) studied the ESR spectra from the chars of a number of polymers and found a correlation between d.c. resistivity and the free electron spin concentration but no correlation with C/H ratio. The g-values they measured were observed to be quite close to that of the free electron, i.e. 2.0026 for cellulosic char vs 2.0023 for the free electron. They concluded that it was not possible to determine whether the observed free spin was due to a or -n electrons. 

In this work, only a brief study was made. The ESR spectra from BPA-PC and BPC-PC chars consisted of a single line and were quite similar to the signal reported for cellulosic char and to each other (Table IV), except that free spin concentration in the BPA-PC burn char was 15X greater than that from BPC-PC and 2.5X greater than observed in the BPA-PC 600° N2 pyrolysis char. It is too early in this study to make any conclusions about these results other than these chars contain significant concentrations of free radicals. 

Chemisorption of oxygen on char has often been discussed previously in terms of free radical concentration in the char (1.5,6). For cellulose chars Bradbury and Shafizadeh (1) found that free spin concentration reached a sharp maximum at HTT 550°C, coinciding with maximum CSA and drew the obvious conclusion that the extent of CSA was at least partly related to free radical content of the char. However, in subsequent work on cellulose char, DeGroot and Shafizadeh (3) have found that unpaired spin concentration continues to increase up to HTT 700"C. Ihe CSA of the char must therefore depend on factors other than free radical concentration. 

Figure 6. SAXS curves for cellulose char activated at 425°C in oxygen to the indicated mass losses.

Scattering data for a cellulose char sample activated isothermally at 425°C are presented in Figure 6. These data show that this carbon, like the saran char, initially exhibits a significant amount of microporosity, as indicated by the plateau centered at about q = O.lA". The Porod invariants for these data initially increase and then decrease with activation, also like the saran char. The scattering data, corrected by the PI values, are presented in Figure 7, which shows somewhat different behavior than for the saran char. 

Figure 8. SAXS curves for Ca-loaded cellulose char during isothermal activation at 340°C in oxygen as a function of time, fhe total burn-off was 73%.

The Porod invariants for the data in Figure 8 increase monotonically with activation. No maximum was observed as for the saran char and unpromoted cellulose char samples. The... 

The original motivation for studying the thermal properties of cellulosic chars came from our study on bulk cellulose pyrolysis under conditions simulating those existing in a fire. In such a situation, the flame over the surface of the solid supplies heat to the pyrolyzing solid. In our work, the radiative and conductive feedback of heat from the flame to the surface was simulated using radiant heaters. The experiments were carried out in an inert gas environment, to maintain as well-defined a heat transfer environment as possible, free from complications due to actual combustion heat sources. A convective How of the inert gas was used to sweep away volatiles from the vicinity of the surface, and the heat transfer effects of the sweep gas were also taken into account. 

Virgin cellulose pellets and cellulose chars produced in the simulated fire apparatus were both examined. Two different measurements were made. One involved measuring the reflected radiation in the mid-infrared from 2.5 to 25 pro (4000 to 400 cm l). These measurements were performed in a diffuse reflectance cell within an FTIR spectrometer. These experiments revealed some wavelength dependence of refleclivity. Reflectance was also measured in-situ in the simulated fire apparatus, by arranging the samples, a fluxmeter. and the heating lamps such that surface reflection of (he incident radiation... 

The above naturally raises a question as to what exactly is meant by the term char in the context of bulk pyrolysis experiments. The definition can clearly only be operational, and is simply the mass that is left when an apparently steady state condition is achieved, regardless of how incomplete the pyrolysis is or however much carbon has resulted from processes involving secondary carbonization reactions of tars and their precursors. Thus all reports of properties of cellulosic chars must be viewed in this light. 

Figure 2. The heat capacities of virgin cellulose (solid line) and cellulose chars (heavy broken lines). Graphite is also shown for reference (thin dashed line).

Figure 3, Thermal conductivity of virgin cellulose (0.458 g/cc - heavy solid line 0.678 g/cc - thin dashed line 0.928 g/cc thin solid line), cellulose chars (heavy dotted lines) and nitrogen gas (thin dotted line).

The porosity of the cellulose chars was studied using nitrogen adsorption at 77K. The results for the fresh cellulose char and (he char burned off to differing extents in oxygen are shown in Figure 5. 

Figure 5. Nitrogen isotherms at 77 K on cellulose chars. The values on the graph refer to the degree of bumoff in 2% oxygen at 723 K.

Figure 7 shows that the development of microporosity is very significant in the wood chars. Just as it was in the cellulose char. In fact, a very similar pattern of surface area development is seen in the wood chars as in the cellulose chars. 

Figure 6. The nitrogen 77 K isotherms on raw cellulose char (filled circles), pine char (open circles) and oak char (open squares). 

The literature shows a wide variation in values of key thermal properties of cellulosics and their chars. Values of thermal conductivity, heat capacity, enthalpy of pyrolysis and surface emissivity apparently cannot yet be safely taken from one study and applied in another. The values of these properties obtained in this study suggest chat the specific heat capacity of cellulose char varies significantly with temperature between about 1.3 J/g-K at ambient temperature to about 2.5 J/g-K at 800 K. The heat capacity of the raw cellulose is comparable to that of the char at low temperatures, but quite a bit higher at temperatures just below the onset of pyrolysis (around 350 K). 

The thermal conductivity of cellulose chars is a strong function of final sample density. Thus the sample porosity (on a macroscopic scale) plays a key role in... 

The surface absorptivity or cmissivity of cellulose chars cannot be safely assumed to be near unity. Near unity values are found for wavelengths in the mid-infrared, but at the shorter wavelengths characteristic of thermal radiation in combustion environments, the cmissivity may be closer to 0,8. Significant energy balance errors may be made in assuming higher values. 

Sekiguchi Y, Shafizadeh F. (1984) The effect of inorganic additives on the formation, composition, and combustion of cellulosic char. J. Appl. Polym. Sci. 29, 1267-1286. 

This investigation was extended to wood and lignin chars prepared at 400 °C to determine the effect of preexisting aromatic nuclei of lignin in the charring reactions. The permanganate oxidation analysis indicated that these chars, like cellulose chars, have considerably condensed or cross-linked aromatic structures, even at 400 C. The NMR data also showed that the chars from similar cellulose, wood. 

Figure 21. Differential scanning calorimetry and thermogravimetry of oxygen chemisorption on cellulose char at 118 C. The analysis was carried out on 2.5-mg samples in aluminum pans using a Cohn R-lOO electrobalance and a DuPont calorimeter cell attached to a DuPont model 990 thermal analyzer, and nitrogen and oxygen gas flows (60 mL/min, dried by passing through H2SO4) were rapidly interchangeable for DSC and TG.

Figure 22, Differential heat of chemisorption as a function of the amount of oxygen adsorbed on cellulose char at 118 °C.

The surface areas of chars prepared from cellulose samples at different HTTs were determined by application of the Dubinin-Po-lany equation to CO2 adsorption at room temperature and compared with the area occupied by surface oxides calculated from oxygen chemisorption at 230 C. The results shown in Figure 25 indicate that cellulosic chars have large surface areas that vary according to the HTT, and peak at about 550 °C. The surface oxides formed by chemisorption occupy only a portion of the total surface area, and the chemisorption also shows a peak for chars formed at about 550 °C, corresponding to the temperature of smoldering combustion. 

The various functional groups were also identified by independent, chemical methods used in organic chemistry [38, 42]. Infrared spectroscopy has also been used for the identification of surface groups. In the beginning, the method suffered from the strong absorption of carbon materials, and poor spectra were obtained. Zawadzki used thin films of cellulose carbonized at 600° C to get acceptable transmission spectra [50], but it is debatable whether such chars are really representative for carbons. It was not possible to heat the cellulose chars... 

Table VI summarizes elemental analyses of chars formed under various pyrolysis conditions. The elemental analyses of cellulosic chars suggest that the composition of chars is not strongly influenced by either the heating rate or sample weight.

---