https://doi.org/10.1021/acsomega.2c07086
“Minimizing the fiber property distribution would have the potential to improve the pulp properties and the process efficiency of chemimechanical pulp. To achieve this, it is essential to improve the level of knowledge of how evenly distributed the sulfonate concentration is between the individual chemimechanical pulp fibers. Due to the variation in quality between pulpwood and sawmill chips, as well as the on-chip screening method, it is difficult to develop an impregnation system that ensures the even distribution of sodium sulfite (Na2SO3) impregnation liquid. It is, therefore, crucial to measure the distribution of sulfonate groups within wood chips and fibers on a microscale. Typically, the degree of unevenness, i.e., the amount of fiber sulfonation and softening prior to defibration, is unknown on a microlevel due to excessively robust or complex processing methods. The degree of sulfonation at the fiber level can be determined by measuring the distribution of elemental sulfur and counterions of sulfonate groups, such as sodium or calcium. A miniaturized energy-dispersive X-ray fluorescence (ED-XRF) method has been developed to address this issue, enabling the analysis of sulfur distributions. It is effective enough to be applied to industrial laboratories for further development, i.e., improved image resolution and measurement time.”
”
Principle of XRF
X-ray fluorescence (XRF) is a technique for determining the elemental composition of a material by measuring the emission of characteristic fluorescent X-rays generated by a high-energy X-ray beam. By analyzing the photons emitted, elements can be determined to be present. (35)
Instrumental Setup
In the XRF experiment, the sample was scanned to obtain the spectrum of a single-spot size, and then, the element of an interest peak was extracted and reconstructed by producing a spectrum of the element of interest. Using a collimated X-ray source and spectroscopic detector, this micro-X-ray technology can produce an elemental spectrum of sulfur and possibly other counterion such as sodium across wood chips or individual fibers in paper samples. An image of our XRF-based miniature setup is shown in Figure 1. To maintain the helium gas environment, a titanium cover palate was affixed to the titanium shield box. There was a 1.6 mm thick titanium plate with a 99.2% purity (metal basis). A two-dimensional (2D) stepper motor unit (Thorlabs) was connected to the sample box. To scan the sample in two dimensions, the stepper motor traveled a distance of 25 mm, and the minimum step size was 0.05 μm. An X-ray tube (Moxtek, 60 kV 12 W MAGPRO) and a spectrometer (Amptek X-123SDD) were mounted in air. The entire setup was covered with lead (Pb) sheets (thickness 1 mm, supplied by Nuclear Shields) into stainless-steel boxes to prevent radiation leakage.
Figure 1
Figure 1. XRF measurement setup before being placed into a stainless-steel box.
The figure shows (a) a MOXTEK X-ray source, (b) an AMETEK spectrometer used to measure the spectrum of elements, (c) a 10 × 10 × 2.5 cm3 titanium shield box to hold the sample at an helium gas atmosphere, (d) helium gas connection, and (e) a stepper motor 2D (XY directional movement).
The focal spot of an X-ray tube is typically 400 μm, and the beam diverges from its source. As a result, a large area of the sample was being used, and the fluorescence photons were accumulating in the spectrometer. To improve the XRF scanning image resolution, the X-ray was collimated using a pinhole to reduce the focus spot size. To obtain a scanning image of individual fibers, the beamline focal point needs to be as small as a few micrometers. An XRF setup was carried out with three-pinhole collimators. The two handmade pinholes had diameters of 10 and 50 μm made of tungsten carbide (WC, Alfa Aesar, 99.95% pure metal) that had shape defects in comparison to a commercial pinhole with a diameter of 100 μm made of gold (Au) and platinum (Pt). A pinhole reduces the beam intensity and increases the measurement time since only a portion of the X-ray passes through the collimator. Spot-scan imaging was employed using two-dimensional precision translation stages, each with a 25 mm travel distance (Thorlabs).
Sample Preparation
Valmet’s CTMP-712 pilot unbleached washed pulp was chosen for sulfur distribution to validate the XRF method. As the CTMP was thoroughly washed, there should be no other sulfur except for covalently bonded sulfur. The strategy was to estimate the differences in sulfur photon counts at high and low proportions of CTMP pulp from different handsheets. CTMP was diluted with bleached softwood kraft pulp (BSWK) of SCA reference kraft K44. An unbeaten low grammage handsheet of 20 gm/m2 was measured for single line scanning in this setup. The low grammage handsheet 20 gm/m2 was prepared by mixing CTMP and BSWK at different percentages according to the ISO 5269-2:2004 standard with tap water using a conventional sheet former with a surface area of 0.021 m2 at the SCA R&D Centre in Sundsvall, Sweden. (36) For the study of sulfonation degree, four different proportions as 100% CTMP, 50% CTMP + 50% BSWK, 30% CTMP + 70% BSWK, and 20% CTMP + 80% BSWK were used.
Earlier studies reported a significantly lower sulfur and sodium circulation system in unbleached kraft pulp, and after bleaching, the total sulfur output and input were the same as those detected by the WinMOPS system in the Metso paper mill. (37) Since unbleached kraft (UBK) pulps with some lignin are thoroughly washed, bleached softwood kraft (BSWK) should not contain any residual sulfur unlike unbleached kraft pulps. It is also expected that the counterion sodium concentration would be too low to detect as well for kraft pulp. Therefore, bleached kraft pulp was mixed with unbleached CTMP pulp at different percentages to dilute the CTMP’s sulfur content. However, fewer sulfur photons were expected gradually by decreasing the percentage of CTMP pulp consistency in these handsheets for the experiment.
Considering the measurement time, a higher concentration of sulfur and sodium was necessary to detect fluorescence photons from light elements. In contrast to sulfur, S (Kα1 2.31 keV), sodium, Na (Kα1 1.04 keV), was challenging to detect with the CTMP handsheet because of its low fluorescence yield. Seltin salt was used as an additional sample for the setup validation since it contains a high concentration of sodium and sulfur compared with other elements. Seltin salt, which is manufactured by Cederroth International AB, contains 50 g of NaCl (per 100 g) as well as KCl, MgSO4, and I. (38)
Results and Discussion
ARTICLE SECTIONS
X-ray Attenuation
Upon excitation of the XRF, fluorescence photons must pass through the media and reach the spectrometer. Fluorescence X-ray radiation from the light element has relatively low energy (long wavelength), and it is severely attenuated by air. Before the XRF setup, it was simulated using a Monte Carlo-N-particle radiation transport code (MCNP). (39) Using X-ray-oriented programs (XOPs), photon transmission curves in air and helium gas relevant for light elements were plotted and can be seen in Figure 2. The simulation assumes 2 cm thickness of air or helium. For air, it results in a lowering of the sulfur peak at 2.31 keV and probably no sodium peak at all was left at 1.04 keV. Helium gas is therefore necessary for the detection of sodium, while sulfur can also be possible to detect in air although the signal decreases. It is possible to think of a vacuum atmosphere as an ideal situation in which there is no air absorption, but the difficulties associated with the maintenance of moving parts in a vacuum have major implications for the design of the instrument at a laboratory scale. For this application, a helium gas chamber is sufficient for measuring sodium.
Figure 2
Figure 2. Calculated photon transmission in 2 cm air and helium.
In this study, a titanium box in a helium gas environment was designed to minimize the fluorescence photon absorption rather than a vacuum environment. Air environment was also considered here, but shield scatter photons in the air interfered with the measurements.
Validation of Feasibility of XRF Setup
To investigate the imaging system resolution of an XRF setup, it is necessary to choose the right pinhole for light elements with small sample sizes. In the air, a preliminary measurement of 60 μm of aluminum (Al) foil was performed. A sandwich structure is shown in Figure 3a, where the aluminum foil is sandwiched between copper and titanium plates. For primary verification of feasibility in the XRF setup, the 80 μm step size of one-line scanning measurement was used to determine the element distribution of Al, Cu, and Ti. A total of 30 steps were scanned. Using a 100 μm pinhole at an X-ray source, the measurement tube setting was 15 keV for 20 min for each step. The photon counts of the characteristic peaks of Al, Cu, and Ti in all measured spectra were integrated. The element distribution map for Al, Cu, and Ti can be seen in Figure 3b.
Figure 3
Figure 3. (a) Image of the Al foil set between Ti and Cu plate. (b) As shown in the figure, Al, Cu, and Ti elements are distributed across X, which are the scanning steps. For each element, the color bar shows the signal intensity.
A fluorescence peak for aluminum appears at 1.48 keV (Kα1) and 1.56 keV (Kβ1). As the spectrometer has a 125 eV full width at half-maximum (FWHM), it cannot distinguish these aluminum peaks. Therefore, they merge into one peak in the output spectrum. Copper has a characteristic peak at 8.04 keV (Kα1), and Ti has a peak at 4.51 keV (Kα1), which is heavier than Al. A higher photon count corresponds to a higher concentration in the color bar. To draw elemental maps, we, therefore, combined spectral information with representative positions. As a result, the maps are in agreement with the sample structure. Due to the air absorption of the Al signal and the low fluorescence yield, the order of magnitude of photon counts from aluminum is much lower than those from titanium and copper.
To test the imaging system resolution of the XRF setup, the histogram of the aluminum signal with Gaussian fitting is shown in Figure 4. According to XRF images, for a 60 μm aluminum foil, the full width at half-maximum (FWHM) was 221.8 μm due to the pinhole size and magnification. Using an 80 μm scan step size, overlapped scanning was used. In scanning measurements, the spatial resolution of the XRF image is limited by the spot size of the source and the scanning step size. In this setup, a pinhole is used to collimate the size of the source beam. As a result of the geometry magnification and pinhole diameter, the size of the focused spot became relatively large for our application. The size of the focused spot can be reduced by reducing the pinhole to a sample distance while keeping the pinhole diameter, by introducing polycapillary optics or by scanning with overlap. A previous study reported the spatial resolution of the XRF image scanning pitch and the mode of scanning (stepwise or continuous) where the 50% overlap (the pitch size is half the focal spot size) is usually considered a possibility to improve spatial resolution. (40,41)
Figure 4
Figure 4. Investigation of the imaging system of the XRF setup using Gaussian fitting of the Al signal.
“