Wood modification with graphene oxide can give it unique features characteristic of other materials. However, the durability of the newly acquired features is of great importance.
To better understand them, it is worth conducting an in-depth analysis of the structural changes that occur in wood under the influence of modification with graphene oxide.
As part of the research, wood was impregnated with aqueous graphene oxide dispersion. Wood was impregnated using two methods: single vacuum and pressureless with ultrasound.
Laser-assisted ionisation spectroscopy (LIBS) was used to determine elements, mainly carbon, and to characterise differences in the elemental composition between the surface layers of wood impregnated with graphene oxide and native wood.
Changes in the structure of polymers building wood tissue were analysed using LIBS and fourier transform infrared spectroscopy (FTIR) spectrometry. The wood surface was also imaged using three microscopic techniques (stereomicroscope, confocal laser scanning microscope, and scanning electron microscopy).
Concept and Function of LIBS Method
Laser-induced breakdown spectroscopy (LIBS) is a microdestructive method used to determine the spatial distribution of elements on the measured surface. This method guarantees quick and precise determination of the qualitative and quantitative composition of elements on the two-dimensional surface of the tested material without prior specialist preparation of the test sample. Another advantage of the method is the possibility of conducting analyses on practically all objects, whether solids, liquids, or gases.
A laser pulse interacting with matter generates plasma, which emits radiation characteristic and at the same time individual for each element, identified as emission lines. The LIBS method helps characterise various materials, including wood.
Cui et al. proposed using LIBS spectroscopy in combination with an artificial neural network (ANN) for species identification and classification of wood. The authors of the study indicated that the distinction of wood such as African rosewood, Brazilian bubinga, elm, larch, Myanmar padauk, Pterocarpus erinaceus, poplar, and sycamore can be carried out based on classification models based on spectral lines obtained by laser spectroscopy methods and computational models established based on training set data.
Other studies have shown that combining LIBS and an electronic nose technique effectively identifies wood species.
LIBS is also valuable for quality control of proper wood impregnation. Solo-Gabriele et al. found that LIBS enables the identification of the impregnation in wood regardless of the quality of the wood, e.g., the presence of knots or water in the wood cells and the method of its finishing, e.g., the use of varnish or paint coatings.
In other studies conducted by López et al., the usefulness of the method was determined for the elemental characterisation of layers in polychromes on wood, which provides extensive knowledge on the methods of renovation of monuments carried out at the turn of the century.
Because laser pulses travel layer by layer, it is possible to profile the depth of individual elements in the layer. Identifying differences in the spectral signals of individual elements at different depths of the material layer enables determining material modification’s structural and molecular mechanisms.
Nasiłowska et al., modifying threads made of stainless steel with graphene oxide (GO), observed explicit incorporation of carbon originating from GO into the surface layer of the native material.
LIBS enabled the authors of the study to observe differences in the spectral peaks of carbon, not only in the surface layer itself and the native material but also in the intermediate layer connecting both layers.
Graphene and its derivatives are of great interest to the scientific community due to their unique properties. As indicated by numerous studies, graphene and graphene oxide have excellent potential in designing functional composites that are useful for many industries.
Moreover, as indicated by the latest research, graphene oxide can be helpful in wood modification to improve its properties or give it new ones, which are characteristic of other engineering materials. Boudjellal et al. presented the results of wood modification with graphene oxide, indicating the modified material’s obtained antistatic properties.
In other studies conducted by Łukawski et al., it was found that graphene materials enables obtaining a superhydrophobic wood surface. Yan et al. showed that wood modification with graphene oxide hinders wood combustion. The researchers also proved that graphene oxide creates a physical and chemical barrier during combustion, stopping the reaction chain.
Significance of Wood Modification
Wood is an environmentally friendly material with huge production possibilities. Wood is suitable as a construction material in producing furniture, tools, laminates, or wood-based panels. Wood has excellent physical and mechanical properties and high durability under low exposure to variable temperature and humidity conditions.
However, the impact of variable humidity on wood is enormous. Therefore, the condition for maintaining the durability of wood in the planned cycle of use is impregnation with protective agents containing biocides.
Wood is also a material that requires fire protection, especially in building structures. Modifying wood with graphene oxide can be a helpful solution in the broadly understood wood technology.
There are various methods of wood modification, the most common of which are chemical impregnation methods.
Depending on the expected conditions of wood use, two methods of wood impregnation affect the depth of impregnation penetration into the wood.
Pressure impregnation enables the saturation of the entire volume of wood with chemicals, while pressureless impregnation is used to modify the surface layers. The way graphene oxide penetrates the anatomical structures of wood may differ from how substances not in the nanoscale penetrate.
Moreover, the way chemicals penetrate wood depends on the moisture content of the wood and the chemical nature of a given substance.
Undoubtedly, the anatomical structure of wood influences the way chemicals penetrate and are located in wood.
Further, the methods and the parameters selected for introducing chemical substances have a great significance on the quality of impregnated wood. The action of vacuum conditions and ultrasound can cause different interactions of chemical substances with wood cells.
In order for graphene oxide to become a commonly used material for wood modification in the future, it is necessary to first learn about the interactions in the wood structure and the most effective ways of introducing it into the wood in such a way as to obtain a composite with the desired properties.
The studies provided valuable information on the structural phenomena occurring during wood modification with graphene oxide.
Wood is a unique technical and technological material that can be modified to improve its properties. However, there are few reports on using graphene oxide for wood modification.
At the same time, the few works that have been published indicate that graphene oxide and its derivatives can give wood unique features characteristic of other materials.
Łukawski et al. showed that graphene wood surfaces can meet the conditions characteristic of electronic requirements.
Graphene materials can also give wood superhydrophobic and anti-corrosion properties. However, the durability of the newly acquired features is of great importance.
In order to better understand these features, an in-depth analysis of the structural changes that occur in wood due to modification with graphene oxide should be carried out.
In this work, the LIBS method was used to determine the elements, mainly carbon, and to characterise the differences in the elemental composition between the surface layers of wood impregnated with graphene oxide and native wood.
This study aimed to determine whether carbon from graphene oxide is incorporated into the surface layers of wood tissue. Identification of this phenomenon was possible thanks to appropriately selected method parameters, which enabled the detection of intermediate layers between the graphene oxide layer and the native material.
In order to confirm the validity of the conclusions, studies were performed using FTIR spectrometry, which confirmed the occurrence of changes in the structure of polymers building the wood tissue and an increase in cellulose crystallinity.
Materials and Techniques Used During the Research Analysis
Pine (Pinus sylvestris L.) samples with dimensions of 40 × 40 × 4 mm were used for the tests. Wood sapwood was used for the tests. Wood density at 12 percent moisture content was 600 kg/m3.
A water dispersion of graphene oxide (Advanced Graphene Products S.A., Zielona Góra, Poland) was used to impregnate wood, diluted with water, to obtain a solution with a concentration of 0.08 percent.
The diameter of the graphene oxide flake was 500–1000 nm. The graphene oxide suspension in water was stirred for 15 min using PS-40A ultrasound to reduce the aggregating graphene oxide flakes in the solution.
The samples were prepared for impregnation by placing them vertically in the solution, one above the other.
They were loaded with glass beads to prevent the wood samples from floating to the surface. Wood impregnation was performed using a single vacuum and an ultrasound. Vacuum impregnation was performed in a vacuum dryer model 1445-2 (with a pump model V-700. The impregnation was carried out under the following conditions: vacuum 190 mbar, time 30 min, temperature 28 deg C.
Then, the pressure in the chamber was brought to atmospheric pressure, and the samples were left in the solution for another 30 minutes. The samples were removed from the graphene oxide dispersion and left to dry.
Pressureless impregnation was performed using an ultrasonic bath. The impregnation process was carried out under the following conditions: time 30 min, temperature 22 deg C. After completing the process, the samples were left in the solution for another 30 minutes. The samples were removed from the graphene oxide dispersion and left to dry.
The wood samples were coded as follows:
•C—Control wood (not treated),
•GO—Wood impregnated with GO using the single-vacuum method, and
•GOU—Wood impregnated with GO using the ultrasonic method.
LIBS studies were performed based on the experimental setup and research procedure presented in the publication.
The studies used a pulsed Nd:YAG laser. The wavelength was 1064 nm, the pulse energy was 10 mJ, and the pulse duration was 4 ns. Plasma radiation was recorded using an echelle ESA 4000 spectrometer.
The spectrometer enabled the detection of spectra in time windows from 20 ns to 16 ms in the 200–800 nm range. The detector was a Kodak KAF 1001 CCD matrix with an ICCD amplifier. The spectral resolution of the entire detection system was λ/Δλ~20,000.
The analysis of stratigraphic spectra consisted of conducting an increasing number of laser shots, which were delivered in the same place, and then recording the number of spectra created by subsequent laser shots.
The focal length of the focusing lens was 150 mm. The distance of the lens from the wood was set to 140 mm to avoid an accidental laser breakdown in air.
The geometry of the sample irradiation was as follows: the laser beam was perpendicular to the wood sample surface, i.e., the angle of incidence calculated from the normal to the surface was 0 degree, while the spectra were collected at the angle of approximately 60 degrees to the surface, i.e., 30 degrees from the normal to the surface.
The diameter of the laser spot focal point on the wood surface with the assumed dimensions of the distance between the lens and the wood was approximately 0.1–0.3 mm, depending on wood age (late or early one). The laser fluence was 0.14 J/mm2. The ablation rate was between 0.08 ÷ 0.12 mm3/pulse and 0.71 ÷ 1.06 mm3/pulse for ablation craters of 10–15 µm in depth.
Since there are visible divisions between early and late wood (annual growth) in the anatomical structure of pine wood, the profile of element spectra was analysed in the early and late wood layers.
Due to the differences in the cellular structure in the early and late wood zones, there is a probability of quantitative and qualitative differences in the graphene oxide incorporated into the wood layers of the zones produced in the spring (earlywood) and summer (latewood) periods.
In order to identify wood samples in the LIBS study, wood samples were provided with codes that also enabled distinguishing the early and late wood zones:
•cb—Control, earlywood zone,
•cc—Control, latewood zone,
•GOb, GOUb—Impregnated wood, earlywood zone, and
•GOc, GOUc—Impregnated wood, latewood zone.
LIBS spectra were treated using multivariate factorial analysis (FA). LIBS uses an orthogonal, linear conversion of the input data (the set of LIBS spectra) in the new variables set called factors.
A Statistica v.10 PL software package was used to process the spectra. The method enabled showing similarities and differences in spectra recorded in different depths of wood samples.
This can be presented in a simple 2D graphical form. In general, the interpretation is the following: the closer the points representing individual LIBS spectra in the graph are, the stronger the similarity of the chemical content in the point from which the LIBS spectrum was created. In contrast, if the spectra are far from each other, they exhibit very differing chemical content.
The sample surfaces were imaged using three microscopic techniques.
Structural studies of the surface of wood impregnated with graphene oxide were performed using:
•A Zeiss Smart-Zoom 5 stereomicroscope. A PalnApo D 1.6×/0.1 FWD 36 mm objective lens with LED illumination was used with a magnification of 35×.
•A LSM 800 Zeiss confocal laser scanning microscope. Analysis parameters: Laser 405 nm, pinhole 0.3 AU, gain 478.
•A scanning electron microscopy (SEM). Images were acquired using an ETD-BSE backscatter detector. The studies were performed at an accelerating voltage of 10 kV. Before imaging, the wood samples were coated with a 6.16 nm thick gold layer. For this purpose, a sputtering device model EM ACE 600 was used. During the sputtering process, the table on which the samples were placed was rotated and tilted at an angle of 120 degrees.
FTIR spectra were recorded using the total internal reflection (ATR) technique in the range of 400–4000 cm−1, at a resolution of 4 cm−1 and a scan count of 64 on a Thermo Scientific spectrometer. The spectra were baseline corrected using Omnic 9 Software.
The absorbances of the peaks taken for calculating the crystallinity indices (TCI, LOI, and HBI) were determined using a local baseline connecting the minima of the corresponding curves from the ATR spectra. The crystallinity index (CI) was determined using three parameters:
•The Lateral Order Index (LOI) from the ratios of the absorbance values at 1430 and 894 cm−1
•The Hydrogen Bond Intensity (HBI) was determined for the ratio of peaks at 3350 and 1337 cm−1
•The Total Crystallinity Index (TCI) for 1373 and 2900 cm−1
Analysis of the Statistical Results of LIBS & FTIR
Example LIBS spectra of the entire recorded spectral range are presented. Selected stratigraphic spectra (pulse numbers 1 and 20) of carbon originating from wood structures and graphene oxide delivered in the surface layers of wood are also presented. These spectra were related to the stratigraphic spectra of the non-impregnated wood surface.
The intensity of 247 nm changes with the pulse number, i.e., depth in the wood, while the differences in the spectrum’s quality result from the wood’s non-uniform structure in the annual growth.
Pine is a species characterised by relatively uniform annual growth. However, it should always be considered that the annual growth width is not ideal. The laser beam focus in deeper layers of wood can go beyond the early wood zone and catch the signal from the second zone—latewood.
The spectra identified in the plasma radiation generated by the laser beam were dominated by carbon compounds and impurities containing the elements—magnesium and oxygen.
In the early control wood (cb), the intensity of the carbon signal measured during the first four laser pulses was low.
In the wood impregnated with graphene oxide, using the vacuum method (GOb), in this wood zone, the signal intensity measured during the first four laser pulses was between 58,087 and 83,916.
Large internal spaces and numerous funnel-shaped cavities on the wall surface characterise the structure of the conducting cells in the early wood zone. These cells efficiently conduct liquids during impregnation processes. Additionally, in vacuum impregnation, impregnants can be sucked deep into these cells.
The latewood zone (cc, GOc) is impermeable to liquids, but the wood cells that make up this zone are characterised by thick cell walls.
In this wood zone, the carbon content from structural polymers, such as cellulose, hemicelluloses, or lignin, should be higher than in the earlywood zone, which is especially visible in the control wood.
The wood cell wall also contains approximately 66 percent carbon. However, it should be noted that numerous phloem rays are composed of parenchyma cells containing carbon compounds in the earlywood and latewood zones. The higher intensity of carbon signals in the deeper layers of the cb zone than the cc zone may result from identifying carbon from cell wall polymers and carbon compounds found in parenchyma cells.
In the near-surface layers of the earlywood zone (GOb) subjected to vacuum impregnation, the higher intensity of carbon signals compared to the latewood zone (GOc) may result from the stronger interaction of GO with cells in this wood zone.
In the earlywood zone, GO can more easily penetrate the conducting cells’ lumen (coils). As shown in previous studies, graphene oxide also penetrates the parenchyma cells and resin ducts, which are numerous in pine wood, usually on the border between earlywood and latewood.
Different results were obtained in the pressureless impregnation method using ultrasound than in the vacuum impregnation method.
Also, about the control, the results of these studies are fascinating. The carbon signal intensity measured in the near-surface layers of the GOUb zone ranged from 3294 to 17,178.
In short-term pressureless wood impregnation methods, the agent’s penetration depth may be small. Nevertheless, the penetration parameters depend on many other factors, such as the wood species, its moisture content, and the chemical nature of the impregnant.
Low carbon signals in the entire wood zone (GOUb, GOUc) may indicate that GO was not introduced into the interior of the wood cells.
Additionally, referring to the results obtained from the tests on control wood samples, it can be assumed that the ultrasound action causes changes in the near-surface layers of wood.
Ultrasound can destroy the parenchyma cells and cause the ejection of cellular substances outside the cell. Ultrasound can also cause cellulose destruction.
Therefore, it can be assumed that the low carbon intensity for samples of wood impregnated without pressure in ultrasound results from the small presence of GO on the wood surface and from the release of wood compounds that are not bound to the matter by physisorption.
Graphene oxide in the early and late wood zones (GOUb and GOUc) can be present on the wood surface, but the action of ultrasound can lead to a decrease in the adhesion of GO to the surface. The zeta potentials of graphene oxide generated in an aqueous environment and during the action of ultrasound can be of great importance in the adhesion of GO to the matter.
The intensity of the oxygen signal in the control wood in the near-surface layers of the earlywood cell zone (cb) is low but increases with depth. A high oxygen signal is observed in the near-surface layers of latewood cells.
It is assumed that this is the oxygen signal originating from the structural components of wood. In the near-surface layers of vacuum-impregnated wood, the intensity of the oxygen signal in the GOb zone is higher than in the GOc zone, which may indicate the appearance of oxygen-containing substances in the near-surface layers.
The intensity level of the oxygen signal in the pressureless ultrasonic-impregnated wood is distributed in a similar way as observed for coal. In the near-surface layers of earlywood (GOUb) and latewood (GOUc), the intensity of the recorded signal is from 319 to 2481 and decreases with depth.
The magnesium content in the near-surface layers of the early wood zone (cb) of the control samples is lower than in the latewood zone (cc), which seems to be correct. The content of mineral components in wood is up to 1.2 percent.
Macroelements such as magnesium and calcium play important roles in cellular processes.
In wood subjected to vacuum impregnation, a higher intensity of the magnesium signal is observed in the near-surface layers (1–5 laser pulses) in the GOb cell zone compared to the GOc cell zone.
The higher intensity of the magnesium signal in the near-surface layers compared to the control results may indicate its presence in the water used for testing. In wood subjected to ultrasound, the intensity of the magnesium signal in the GOUb zone is similar to the results obtained for the GOb zone.
In turn, the intensity of signals in the surface layers of the GOUc zone (1–3 laser pulses) is very high. It can be assumed that this signal comes from impurities (magnesium ions), the source of which is the water used for testing.
Because a heterogeneous anatomical structure characterises wood, the zone of early (spring) wood is anatomically different from the zone of late (summer) wood; the structure of the wood in the annual growth ring can significantly affect how wood tissue interacts with substances and materials introduced into it.
Early wood cells are characterised by thinner cell walls and larger cell lumens, which enables them to contain a more significant number of graphene oxide flakes. Thicker cell walls characterise latewood cells but are practically impermeable.
The relationship between the relative number of elements C, Mg, and O in wood was confirmed and the different structures of the wood in spring and summer growths.
The signal value (counts) mean the area under spectral line of the interest (C 247 nm, Mg 279 nm, and O 777 nm, respectively) corrected for the background contribution.
In the control wood, the content of elements in the near-surface layers, primarily carbon, is higher in latewood, which is a regularity resulting from the more significant share of lignocellulose in the cell walls. In the case of wood impregnated with graphene oxide using the vacuum method, the amount of the elements tested in the near-surface layers, primarily carbon, was dominant early, which indicates that large and empty spaces of the cells that make up the wood were filled with GO flakes.
The carbon, magnesium, and oxygen content in the near-surface layers of wood impregnated with graphene oxide using the ultrasonic method is lower than that of vacuum-impregnated wood.
Results of the LIBS spectra factorial analysis (FA) are presented. It is seen that there are small differences between spectra taken in bright and dark wood grains which show similar chemical composition.
The location of the spectra from GOb, GOc, GOUb, GOUc, as well as from control not impregnated samples (cb and cc) also demonstrate certain differences in chemical content.
Structural studies conducted using a scanning electron microscope of impregnated wood and pressureless ultrasound indicate the presence of graphene oxide flakes on the wood surface.
Graphene oxide flakes form clusters and surface undulations. Single GO flakes are visible in the microscopic image. The surface of untreated wood is flat without artifacts.
The presence of graphene oxide flakes on the surface of impregnated wood was also visualised using a stereoscopic microscope. The early and late wood zones became darker. Graphene oxide changes the natural colour of the wood surface, causing it to turn grey. Small clusters of deposited GO are visible on the surface of impregnated wood, shown as dark dots (marked with a black arrow).
As indicated by the research conducted by López et al., LIBS can be used to clean the wood surface from layers of applied coatings.
The wood structure imaged with a confocal microscope indicates changes occurring on the wood surface due to laser ablation.
The depth of the crater (damaged wood surface) created due to the impact of 20 laser beam pulses locally located on the surface of the samples was more significant in the early wood zone than in the latewood zone.
This means thinner layers of the wood cell wall in the early wood zone are more susceptible to damage than the cell walls in the latewood zone. Having in mind that crater diameters were typically approximately 100 µm for latewood samples and approximately 300 µm for early ones, we can find that after 20 laser shots the averaged total (effective) ablation volumes were 0.4 mm3 for a late pine wood and approximately 8.8 mm3 for early one (we assumed a cylinder-shape crater in calculations).
The FTIR spectra obtained as a result of the measurement were compared to show the changes that occurred due to the in laying of wood with GO.
Due to the high cellulose content in wood, the analysis of structural changes in wood after impregnation with graphene oxide was related to cellulose parameters. The areas (peaks) associated with the crystalline and amorphous structure were compared; on this basis, the influence of the applied GO on the wood structure was determined.
The presence of the amorphous form of cellulose is determined based on the 2900 cm−1 peak (C-H stretching bonds), its size, and position after the modification was introduced.
An increase in amorphousness was observed when the peak intensity decreased and its maximum shifted towards longer wavelengths. The second amorphous band is the 898 cm−1 which is associated with the β-glycosidic bond, assigned to the C-O-C stretching vibrations.
The 1430 cm−1 band, associated with the symmetric -CH2 bending vibrations, is associated with the crystalline part of cellulose. The decrease in the band intensity is corelated with the decrease in cellulose crystallinity.
The spectrum of examined samples shows characteristic peaks for cellulose compounds. The range of 3200–3500 cm−1 corresponds to -OH stretching vibrations, the bands around 2930–2840 cm−1 are related to C-H stretching vibrations, and the peaks 1732, 1638, 1420, and 1370 cm−1 are related to the presence of C = O stretching, -OH bending, C-OH stretching and C-H bending bonds, respectively.
The peaks 1154 and 1055 cm−1 are related to the cellulose skeleton in the form of C-O-C and C-O bonds. No new peaks are observed after GO is introduced to the surface of the control sample for both variants (GO and GOU).
A slight increase in absorbance was recorded for the GOU sample in the area of 2500–3600 cm−1. Due to the lack of changes in other areas, it can be assumed that these changes are the results of GO layer deposition on the wood surface (influence of -OH and -CH2 groups).
The LOI parameter represents the regions ordered in the cellulose, in this case it may indicate that the GO flakes are arranged in a direction perpendicular to the cellulose chain. The HBI parameter is unchanged for all samples, which indicates that no hydrogen bonds have been formed after the GO layer has been applied.
The surface modification was carried out at 22/28 deg C, so these were mild conditions for the pine substrate (control sample). Such deposition methods may not affect the surface significantly. The TCI parameter increased after GO deposition, but since the IR beam penetrates the surface at 5–15 µm in depth, the change is strictly local.
Conclusions of the Research
This study used the LIBS method to identify mainly carbon and oxygen in wood impregnated with graphene oxide. The method used enabled us to characterise differences in the signal intensity originating from the elements depending on the wood impregnation method.
A higher intensity of the signal originating from carbon and oxygen was identified in the near-surface layers in the early wood cell zone. The low intensity of the signals of the tested elements on the surface of wood impregnated using the pressureless ultrasound method may indicate disintegration processes occurring in the near-surface cell structures of wood.
Despite the low signal intensities recorded in the LIBS study for wood impregnated using the pressureless method, a slight increase in absorbance was noted in the range of 2500–3600 cm−1 in the FTIR studies.
It can be assumed that these changes result from the GO layer deposition on the wood surface (the effect of the -OH and -CH2 groups). Microscopic examinations indicate the presence of graphene oxide flakes in the cells of the surface layers of wood.
Based on the studies, the LIBS technique correlated with the FTIR method can be used to identify changes occurring in the surface layers of wood subjected to various impregnation methods.
