PAPERmaking! Vol5 Nr2 2019

Cellulose

mode with a secondary electron detector was used for imaging samples. It was necessary to coat samples with a thin layer of gold in order to obtain quality SEM images.

was determined in cupri-ethylenediamine solution (CED) according to ISO 5351-1 (1981).

Chemical analysis

FT-NIR measurement

The concentration of cellulose was determined according to the Seifert procedure by using acetylace- tone-dioxane-hydrochloric acid (Browning 1967). The quantities of other wood components were determined according to the following standards: • Hot water extractives (T207 cm-08 TAPPI 2008). • 1% NaOH extractives (T212 om-07 TAPPI 2007). The solubility in 1% H 2 SO 4 was performed following the same procedure as in 1% NaOH (T212 om-07 TAPPI 2007). All the chemical analyses were per- formed before and after degradation and were repeated three times and the maximum standard deviation of results was considered as the indicator of the mea- surement error.

NIR measurements were made using a Vector 22-N spectrophotometer (Bruker Optics GmbH, Ettlingen, Germany). The measured spectral range was between 4000 and 12,000 cm - 1 (2500 nm and 833 nm) with a resolution of 8 cm - 1 . The spectral wavenumber interval was 3.85 cm - 1 with zero-filling = 2. An average of 32 consecutive measurements (internal scans) was acquired for each spectrum. The degrada- tion stage of all papers was determined on three representative paper strips and measured on three spots, before and after 4 and 8 weeks of degradation in each soil. The resulting spectra (9 spectra) were averaged and stored in a data base for further analysis. All measurements were performed in an air-condi- tioned room (20  C and 65% RH) in order to minimize the effects of the temperature and moisture variations.

Statistical data analysis

A factorial one-way ANOVA was conducted to compare mean values of chemical components con- centration and paper properties at significance level p \ 0.05. In addition, Tukey contrasts simultaneous test for general linear hypothesis with multiple com- parison of means was performed post hoc to identify groups of statistically similar samples. The analysis were performed in R software (www.r-project.org) by using multcomp module.

Spectral data analysis

The interpretation of spectra based on Schwanninger et al. (2011) and corresponding band assignments are summarized in Table 2. A narrow range, between 4000 and 6000 cm - 1 , was selected for spectra inter- pretation. Computation of the second derivative [21- point smoothing, Savitzki and Golay algorithm (1964)] was applied for spectra pre-processing. Spec- tra were also pre-processed with extended multiplica- tive scatter correction (EMSC) algorithm for investigation of the degradation rate. Principal com- ponent analysis (PCA) was used for data analysis. The commercially available OPUS 7.0 (Bruker Optics

Electron microscopy

A scanning electron microscope (SEM; S-3400 N, Hitachi High-Tech, Tokyo, Japan) in high vacuum

Table 2 Interpretation of the FT-NIR molecular vibrations (Schwanninger et al. 2011)

Wavenumber (cm - 1 )

nr

Band assignment

Chemical component

1 C–H def.

Holocellulose

4198

2 C–H def. ? C–H str.

Cellulose

4280

3 C–H 2 def. ? C–H 2 str.

Cellulose

4404

4 C–H def. ? O–H str.

Cellulose

4620

5 C–H def. ? O–H str.

Cellulose

4890

6 O–H def. ? O–H str.

Water

5219

7 C–H def. ? O–H str.

Semi-crystalline/crystalline cellulose 5464

123

Made with FlippingBook - Online catalogs