For all samples, the measured refractive index *n*(*ν*) is essentially independent of frequency within the relevant frequency range (0.3–0.7 THz). Therefore, it is reasonable to average the value over this entire frequency range, reducing the data to a single averaged value. In addition, the measured extinction coefficients *κ*(*ν*) exhibit a very weak (nearly linear) increasing trend with frequency. This trend is weak enough that it is reasonable to perform a similar spectral averaging. Using these averages, both the refractive index and the extinction coefficient exhibited a consistent trend with respect to the time of fungus infestation: the values of both *n* and *κ* decrease with increasing infestation time, with a few exceptions. The average refractive index of non-infested dry wood samples was determined to be 1.42±0.03. This value decreased to 1.09±0.01 after 20 weeks of infestation. The average extinction coefficient changed from 0.055±0.0065 for non-infested samples to 0.018±0.0024 after 20 weeks of infestation. These extinction coefficient values are much less than unity, which justifies the approximation that *n*<<*κ*, as noted earlier. A similar trend is also observed for the mass of most of the samples. Figure 1 shows the mass loss (in percentage) vs. infestation time.

Figure 1: Mass loss of the samples at different infestation times.

For reference samples (non-infested), the mass loss equaled zero within 2% tolerance (not shown in the plot).

The observed trend of mass loss is not surprising, as fungi cause decomposition of wood as their source of energy, resulting in mass and density reduction (i.e. increased mass loss). The speed at which fungi grow can vary, despite taking care that all samples were prepared in the same way. This fact contributes to the scatter in the data.

In many previous studies, it is empirically observed that both the THz refractive index and extinction coefficient are related to the density of the material (Koch et al. 1998); Naftaly and Miles 2007; Palermo et al. 2008; Laib and Mittleman 2010; May et al. 2013). Thus, it is logical to seek a correlation between the measurements of refractive index and extinction coefficient and the measured mass loss. First, the refractive index loss and extinction coefficient loss are computed as:

$$\text{Relative}\text{\hspace{0.17em}}\text{refractive}\text{\hspace{0.17em}}\text{index}\text{\hspace{0.17em}}\text{loss}=\frac{{n}_{1}-{n}_{2}}{{n}_{1}-1}$$(2)

$$\text{Extinction}\text{\hspace{0.17em}}\text{coefficient}\text{\hspace{0.17em}}\text{loss}=\frac{{\kappa}_{1}-{\kappa}_{2}}{{\kappa}_{1}},$$(3)

where *n*_{1} and *κ*_{1} are the spatially and spectrally averaged refractive index and extinction coefficient of non-infested wood samples, respectively, and *n*_{2} and *κ*_{2} are the averaged refractive index and extinction coefficient of an infested sample, respectively. Note that a value of *n*_{vacuum}=1 is subtracted from *n*_{1} in the denominator of equation (2). If the entire sample was completely degraded by the fungal infestation, the refractive index *n*_{2} would be 1 (equivalent to no sample), which would correspond to 100% mass loss. Figure 2 shows the results, displaying the correlation between refractive index loss and mass loss (Figure 2a) and the correlation between extinction coefficient loss and mass loss (Figure 2b). In both cases, linear function fitted the data well with an R^{2} value of 0.99. Evidently, the results exhibit a very strong correlation in both cases. In addition, a Monte Carlo approach was employed to determine the correlation coefficient between optical parameter loss and mass loss taking also the measurement uncertainties into account. In short, many random datasets normally distributed within the error bars are created and an average correlation coefficient and its standard deviation are calculated. We found a correlation coefficient of 0.88±0.05 for the refractive index and 0.69±0.14 for the extinction coefficient. These results suggest that THz spectroscopy could be a valuable non-contact and non-destructive tool for the assessment of mass loss due to fungal infestation of wood.

Figure 2: Correlation of dielectric parameter loss with mass loss of the samples.

(a) Relative refractive index loss vs. mass loss with a linear fit, (b) extinction coefficient loss vs. mass loss with a linear fit. In both cases the linear function with a fixed intercept at (0,0) was fitted and the coefficient of the determination (R^{2}) was 0.99. Slope was determined as 1.05 and 0.88 in the case of relative refractive index loss and extinction coefficient loss, respectively.

Evaluation of the microscopic analysis: In the common beech wood, three types of degeneration by *T. versicolor* were observed:

round to oval cavities,

fine growth through cell walls,

planar cell wall degradation starting from the lumen, which expands with time.

Observing this biotic degradation confirms that the fungus infestation took place in a foreseen way (Tuor et al. 1995). The results are shown in Figures 3–5. Hyphae growth took place before the degradation of wood components, which began in the vessels and proceeded from there into the parenchyma and fiber cells. The mycelium density was already high by the 4-week mark, increased until the eighth week, stagnated in the 16^{th} week and decreased or remained at the same level at the 20^{th} week. The hyphal diameter increased to 2–4 μm until the eighth week and then stabilized, with the mycelete density in the parenchymal cells and fibers being low and then variable. The colonization was very dynamic and was then partially decomposed autolytically (dissolving of hyphae). When looking at the cutting directions, there were no substantial differences in the degradation patterns. However, the trend shown in the THz analyses and mass loss measurements can be confirmed in the microscopic analyses. In the following, the changes in the individual degradation patterns of the weeks are compared in the microscopic analysis.

Figure 3: The number of round cavities increases significantly over time.

After 4 weeks they are sparse, moderately frequent after 8 weeks, more frequent after 16 weeks and numerous after 20 weeks.

Figure 4: In the fourth week, few caverns and many fine wall penetrations were visible.

These can merge into each other over time.

Figure 5: The first indications of a planar cell wall degradation appear after 8 weeks and become clear after 16 weeks.

After 20 weeks, it was the most dominant feature of the damage picture.

In this comparison, the results for the control samples are not presented.

In the comparison of non-infested beech wood of a control sample and infested material of a sample both from the 20^{th} week, it becomes clear how strongly the degradation of the cell walls took place (Figure 6).

Figure 6: Left side, Control sample junction field and fibrous tissue with brown deposits (red arrow). Right side, Infected wood sample (20 weeks) cell wall rejuvenation by the flat cell wall degradation (green arrows) and round to oval cavities of white-rot fungus.

Visible are numerous hyphae.

The microscopic analysis was performed on a single infested sample from each test group (i.e. infestation time of 4, 8, 16, 20 weeks) and a single reference (non-infested) sample. This means that in total five samples were investigated under a microscope, as the sample preparation for the microscopic investigation as well as investigation on its own is quite time-consuming. The selected samples followed the trend of losing mass and reducing refractive index with increasing infestation time, which was also qualitatively confirmed with the microscopic analysis. The evaluation of the results of the microscopic analysis are summarized in Table 1.

## Comments (0)

General note:By using the comment function on degruyter.com you agree to our Privacy Statement. A respectful treatment of one another is important to us. Therefore we would like to draw your attention to our House Rules.