Introduction

Plantation wood is an ideal material for sustainable society. However, due to the inherent hygroscopicity of wood, the dimensional stability of wood is poor, which limits its application. The dimensional stability of wood can be improved by reducing the hydrophilic –OH group of cell wall components or bulking the cell wall with a modifier. There are many studies on dimensional stability of chemically modified wood. Typical modifications of wood are esterification (Hill and Jones 1999; Kumar 2007), etherification (Chang and Chang 2006), resin impregnation (Gindl and Gupta 2002; Furuno et al. 2004), silylation (Hill et al. 2004; Donath et al. 2004) and in situ grafting polymerization of monomers (Wang et al. 2018; Huang et al. 2019). Although chemical modification can improve wood properties, most treatments involve harmful chemicals or solvents, which may cause serious environmental and health problems in the process of processing and using modified wood.

N-isopropylacrylamide (NIPAM) is a water-soluble vinyl monomer and of low toxicity (Naha et al. 2009; Ghavami et al. 2010). Poly-(N-isopropylacrylamide) [p(NIPAM)] is one of the most widely studied thermoresponsive polymers, with a lower critical solution temperature (LCST) near human body temperature in aqueous solutions. When the water solution temperature of p(NIPAM) is higher than the LCST, p(NIPAM) changes from hydrophilic to hydrophobic (Śliwa et al. 2017; Raju et al. 2018). Cabane et al. (2016) grafted NIPAM onto wood by ATRP to prepare thermosensitive wood. Modified wood has lower dimensional stability at lower temperatures and higher dimensional stability at higher temperatures. Keplinger et al. (2016) synthesized a novel smart hierarchical bio-based material by grafted NIPAM thermosensitive hydrogel onto wood cells in two steps. The temperature dependence of thermosensitive hydrogels was studied by Raman spectroscopy.

2-Hydroxyethyl methacrylate (HEMA) is a water-soluble organic monomer with low toxicity and no irritating smell. P(HEMA) has good biocompatibility and is insoluble in water (Horák et al. 1997; Walther et al. 2004). P(HEMA) is widely used in contact lens manufacturing, dental fillers, tissue engineering scaffolds and drug delivery (Miyasaka and Nakabayashi 1999; Lou et al. 2004; Abraham et al. 2005). Chen-Yang et al. (1998) used HEMA and hexachlorocyclotriphosphazene to prepare reaction intermediates and then used UV radiation to initiate in situ polymerization to improve wood flame retardancy. Cleland et al. (2009) used X-ray to initiate in situ polymerization of HEMA in wood cell wall. The results showed that the polymer could not be extracted from the cell wall.

In this paper, environment-friendly modification methods were used to improve wood dimensional stability. The mechanism of the grafting reaction between vinyl monomers and wood cell wall was studied by X-ray photoelectron spectroscopy (XPS). The results of this study can provide a framework for improving wood water resistance with water-soluble monomers and water as solvents.

Materials and methods

Materials

Poplar wood (Populus euramericana cv.I-214) was obtained from a fast-growing plantation located in Hebei Province, China. 2-Hydroxyethyl methacrylate (HEMA), N-isopropylacrylamide (NIPAM) and 2,2′-azobis[2-methylpropionamidine] dihydrochloride (AAPH) were purchased from Aladdin Industrial Corporation (Shanghai, China). Deionized (DI) water was used as solvent for the preparation of the impregnation solutions.

Chemical modification of wood blocks

The functional monomers HEMA and NIPAM were mixed at a mass ratio of 6/1 and added to deionized (DI) water resulting in an aqueous solution with 40% monomers content. Moreover, 1% of AAPH (by wt. based on the functional monomers) was added and stirred at room temperature for 10 min. Then, a vacuum/pressure treatment was applied for impregnation (− 0.1 MPa for 30 min followed by 0.8 MPa for 120 min) of the experimental blocks. After removal from the solution, the sample surfaces were washed with deionized (DI) water. The impregnated samples were heated at 60 °C for 12 h, followed by heating at 103 ± 2 °C for 12 h. After the polymerization, the modified wood blocks were immersed in aqueous solution for half an hour to wash off the residual modifiers on the surface. The weight percent gain (WPG) was determined by weighing, which was around 43% (the standard deviation was 2.8 according to the weight gain rate of 20 samples).

Characterization

The surface of wood was sliced smoothly by a rotary microtome (LEICA RM2255, Germany). The morphology of the wood was characterized by field-emission scanning electron microscopy (FESEM) (Hitachi S-8010, Japan) equipped with an energy-dispersive X-ray (EDX) detector for mapping (Horiba Scientific, Japan). Fourier transform infrared spectra (FTIR) of samples were recorded on a Nicolet 6700 instrument by the KBr pellet pressing method (Thermo Scientific, USA). X-ray photoelectron spectroscopy (XPS) was performed on a Thermo ESCALAB 250 spectrometer (Thermo Scientific, USA).

Physical properties

Measurements of weight percent gain (WPG), water uptake (WU), mass loss (ML) and antiswelling efficiency (ASE) were taken with 10 replicates of cubic samples with a size of 20 mm × 20 mm × 20 mm (longitudinal × radial × tangential). Samples were submerged in deionized water at 25 °C for 120 h and then dried in an oven at 65 °C until constant mass was obtained. ASE and MS were calculated after measuring the volume and weight of samples before and after immersion. The oven-dry wood blocks were immersed in deionized water for 30 days, weights were measured at intervals and water uptake (WU) was calculated.

The weight percent gain (WPG) was calculated as:

$${\text{WPG }}\left( \% \right) = \left( {{\text{Wt}} - {\text{Wu}}} \right)/{\text{Wu}} \times 100$$

where Wt represents the oven-dried weight of chemically modified sample, Wu represents the oven-dried weight of unmodified sample.

The mass loss was calculated by the following formula:

$${\text{ML }}\left( \% \right) \, = \, \left( {{\text{Wa}} - {\text{Wb}}} \right)/{\text{Wa}} \times 100$$

where Wa represents the oven-dry weight of samples before each cycle of soaking, and Wb is the oven-dry weight of samples after each cycle of soaking.

Water uptake (WU) was calculated as:

$${\text{WU }}\left( \% \right) \, = \, \left( {{\text{Ws}} - {\text{Wd}}} \right)/{\text{Wd}} \times 100$$

where Wd represents the weight of sample before each cycle of soaking, and Ws is the weight of sample after each cycle of soaking.

ASE was calculated from the wet and oven-dried volumes of the treated and untreated wood specimens:

$${\text{ASE }}\left( \% \right) = \left( {{\text{Su}} - {\text{St}}} \right)/{\text{Su}} \times 100\%$$

where Su is the volumetric swelling coefficient of the untreated blocks, and St is the volumetric swelling coefficient of the treated wood blocks.

Volumetric swelling coefficient was calculated as follows:

$$S \left( \% \right) = \left( {V2 - V1} \right)/V2 \times 100\%$$

where V2 is the volume of the water-saturated wood blocks, and V1 is the volume of the oven-dried wood blocks.

Results and discussion

SEM–EDX analyses

Figure 1 shows the SEM images of untreated (A) and treated samples (B). It was observed that in the unmodified wood, there are many gaps between cell walls and the CML. However, in the modified sample, there is no space between cell walls and the CML. By comparison, the average thickness of the modified wood cell wall becomes thicker than before modification. According to the Nano Measure software statistics, the average thickness of the untreated cell walls was 3.10 ± 0.5 μm (a), and the average thickness of the treated wood cell walls was 3.90 ± 0.5 μm (b) (distribution ratio: the ratio of the number of cell walls corresponding to the thickness to the total number of cell walls measured). This could be related to the two monomers penetrated into the cell wall and copolymerized in the cell wall, bulking the cell wall and grafting into cell wall.

Fig. 1
figure 1

SEM images of unmodified wood (A) and modified wood (B). Elemental maps of C, N and O are shown for each class. Element content of modified wood (B-1) and natural wood (A-1). Thickness distribution of untreated cell wall (a) and treated cell wall (b)

Full size image

Columns 2–4 show EDX elemental maps of C, N and O to identify the distribution of the different elements within the samples. The content of nitrogen in the cell wall increased from 0 before modification to 3.6% after modification, indicating that the modifier penetrated into the cell wall. From the distribution of carbon, nitrogen and oxygen in the cell wall of wood before and after modification, it can be seen that the elements are evenly distributed in the cell wall, indicating that the modifier penetrates into the cell wall of wood evenly [element content of modified wood (B-1) and natural wood (A-1)].

FTIR analyses

Figure 2 shows the FTIR spectra of the unmodified wood, modified wood and polymer. The spectrum of grafted wood exhibits main peaks at 1650 and 1542 cm−1 corresponding to the characteristic peaks for amide I and amide II, respectively. The strong and narrow absorption peak shown at 1730 cm−1 (–COO–) confirms the presence of the carbonyl group of ester (Liu et al. 2014; Jin et al. 2018). The absorption bands located at 3400–3500 cm−1 correspond to aromatic and aliphatic hydroxyl groups; 1590 and 1510 cm−1 correspond to aromatic rings (Chang and Chang 2006; Wang et al. 2018). Infrared spectra showed that water-soluble monomers were in situ polymerized in wood.

Fig. 2
figure 2

FTIR spectra of unmodified wood, modified wood and polymer

Full size image

XPS analyses

Wood blocks were ground into wood flour with a particle size of about 300 microns and then studied by XPS. The XPS survey spectra are presented in Fig. 3. The main elements detected are oxygen (532.6 eV), carbon (285.3 eV) and nitrogen (398.9 eV). Nitrogen is present in the form of amide functional groups. Nitrogen content increased from 0.65% before modification to 1.73% after modification.

Fig. 3
figure 3

XPS survey spectrum of unmodified (a) and modified wood (b)

Full size image

According to most data in the literature, the carbon spectrum (C1s) was deconvoluted into four types. C1: carbon atoms bonded only to carbon or hydrogen (C–C, C=C, C–H). C2: carbon atoms bonded to a single noncarbonyl oxygen (C–O, C–O–C). C3: carbon bonded to a carbonyl or two noncarbonyl oxygen (C=O, O–C–O). C4: carbon atoms bonded to a carbonyl and a noncarbonyl oxygen (O–C=O, C(=O)OH) (Bryne et al. 2010; Tuong and Li 2011; Fernández-Fernández et al. 2014).

Figure 4 shows the XPS high-resolution spectra of C1s. Typical BE values for the samples are C1: C–C and C–H at 284.6 eV, C2: C–O and C–O–C at 286.2 eV, C3: C=O and O–C=O at 287.6 eV and C4: O–C=O and C(=O)OH at 288.6 eV. Water-soluble ethylene monomer treatment increases C1 slightly from 28.83 to 32.91%, whereas C2 reduces from 57.10 to 53.35%. In addition, the content of C4 increased from 3.5% before modification to 5.5% after modification, indicating that the monomer was grafted onto the C–O of the wood cell wall.

Fig. 4
figure 4

XPS high-resolution spectra of C1s for modified wood (right) and unmodified wood (left)

Full size image

Improved wood properties

The weight loss of samples during repeated immersion is due to the removal of extracts and unreacted water-soluble vinyl monomers. The mass loss value of the sample after the immersion cycle is shown in Fig. 5a. The extracts of untreated samples were removed after two immersion cycles, but for treated wood samples, most of the unreacted water-soluble vinyl monomers were removed after three immersion cycles.

Fig. 5
figure 5

Mass loss (a), dimensional stability (b) and water uptake behavior (c) of the wood samples

Full size image

After three soaking cycles, the ASE value of modified wood is as high as 50%, which means that the insoluble resin formed by water-soluble monomers can be well anchored on the cell wall of wood, effectively reducing the water absorption of wood, while preventing the swelling of wood due to water absorption (Fig. 5b). Water-soluble monomers of small molecules can enter the cell wall and polymerize in the cell wall. The insoluble polymer formed will swell the cell wall and hinder the entry of water, thus improving the dimensional stability of treated wood.

Figure 5c shows the results of deionized water immersion during 30 days. As visible, both the untreated and treated wood absorb water quickly within the first 5 days. Then, the curves flatten. During the whole observation period, the treated wood absorbs less water than the untreated one. The final water uptake of the treated wood decreased from 163 to 109% when compared to untreated wood. The polymers bulk the cell walls, thereby preventing water from entering the cell walls.

In summary, the graft copolymerization improved the dimensional stability and water uptake behavior.

Mechanism of graft polymerization

Based on the results of the current experiment, the mechanism of the grafting copolymerization between wood and water-soluble vinyl monomers is proposed. As depicted in Fig. 6, the initiator decomposes into free radicals under mild conditions (reaction 1), which are able to abstract hydrogen from wood to form wood macromolecular radicals (reaction 2). It is presumed that once the wood macromolecular radicals are generated, they can support active grafting sites to monomer to anchor onto the wood (reaction 3) (Qiu et al. 1999; Ivanova et al. 2016).

Fig. 6
figure 6

Proposed mechanism of graft copolymerization of water-soluble monomers onto wood using AAPH initiator

Full size image

Conclusion

In this study, water was used as solvent to improve wood dimensional stability by in situ polymerization of water-soluble vinyl monomers. The results showed that the dimensional stability (ASE) of modified wood reached 50%. The results from FTIR and XPS indicate that the free radical reaction could graft vinyl monomers onto wood.