The properties of a paper depend on those of the wood or pulp fibers from which it was produced. For example, individual fibers are known to vary widely even within the same tree. Nowadays, in a context of forest and limited wood supply management, the increasing demand for high-quality paper products requires more efficient use of available fiber resources. Research and analytical tools for measuring the properties of individual pulp fibers are accordingly considered essential to improve use of this resource.
A wood fiber can be described by its middle layer, S2, which contains the majority (80–95%) of the cell wall material.1 The angle between the fibrillar direction and the fiber axis φ is termed the microfibril angle (MFA) (see Figure 1). A number of experimental studies have shown that the MFA is closely related to the mechanical properties of fibers, such as strength, elastic modulus, and shrinkage.2,3
When a fiber is illuminated by polarized light, a relative phase retardation Δ is produced between the two orthogonal components of the light traveling in the S2 layer as it emerges from a cell wall (see Figure 1). The retardation Δ or path difference PD is proportional to the cell wall thickness (CWT) d, with Δ=2πd(n2−n1)/λ or PD=Δλ/2π=d(n2−n1) where (n2−n1) is the birefringence of the wall material and λ is the wavelength of the light. The CWT is related to fiber flexibility, strength, and collapsibility. In applications, it is important to know the distribution of CWTs instead of their absolute values so that Δ or PD can be used to replace CWT. In addition, the CWT can be determined from Δ or PD through calibration.
Figure 1. Schematic representation of the S2 layer of a single wood fiber.
φ is the microfibril angle and Δ is the phase retardation (see text).Both MFA and CWT are difficult to measure for fibers due to their two-walled structure. Recently, a new method based on spectroscopic transmission ellipsometry (STE) was developed to allow quick measurement of pulp fibers oriented arbitrarily for both φ and Δ.4,5 As shown in Figure 2, the STE method uses an optical arrangement consisting of two λ/4 retarders placed between a pair of polarizers. The retarders and polarizers are oriented such that any image generated is insensitive to the orientation of the fibers, relying instead only on their φ and Δ values. An ImSpector device placed after the optical arrangement scans the fibers and captures a line image across the fiber samples: see Figure 3(a). It also disperses light I from the fiber segments selected for measurement into spectrum I[Δ(λ), φ], as shown in Figure 3(b). A background image segment I0—see Figure 3(a)—bordering the measured fibers is selected as a reference that approximately describes the light intensity transmitted by the whole system at the position of a measured fiber segment for regions where a fiber is missing. The reference image segment I0 is dispersed into I0(λ)—see Figure 3(b)—which is used to normalize I[Δ(λ), φ] such that the spectral transmission function T[Δ(λ), φ] of a measured fiber segment can be described by T[Δ(λ), φ]=I[Δ(λ), φ]/I0(λ), which is independent of the spectral properties of the equipment used. Both φ and Δ are then evaluated by fitting T[Δ(λ), φ] to4,5
and comparing the fit to the measured curve using a least-squares procedure (see Figure 4).
Figure 2. Optical spectroscopic transmission ellipsometry arrangement for measuring pulp fibers. The setup consists of two λ/4 retarders (φ1, φ2: orientation angles) arranged between a pair of polarizers (P1, P2: azimuths).
Figure 3. (a) Micrograph of three pulp fibers. Ia, Ib, and Ic: fiber segments selected for measurement. I0: selected reference from the background image. (b) Spectral image dispersed from the scanned image part in (a).
Measured spectral transmission functions T[Δ(λ), φ] of the fibers shown in Figure 3
(a) with best fits in the 405–710nm range.
The advantage of the STE method is that it requires neither sample pretreatment nor fiber alignment. Moreover, only a single image is needed to measure intact pulp fibers with arbitrarily oriented microfibril angles φ and retardation Δ. The next step will be to design STE equipment suitable for use not only in the laboratory but also under on-line conditions. It is anticipated that such equipment would be able to provide the much-needed measurements of φ and Δ as well as those of other parameters such as fiber length, width, and shape.
Measurement and Sensor Laboratory,
University of Oulu