Depth profiling of photodegraded wood surfaces by confocal Raman microscopy

Introduction

Weathering leads to rapid depolymerization of the chemical components and degradation of the cellular structure of wood [1, 2]. These changes are facilitated by a combination of solar radiation, moisture, temperature, air pollutions, and other factors. Among them, solar radiation is most damaging to wood because ultraviolet (UV) rays cause free radical oxidation and the reactions that significantly deteriorate the constituents of wood [1, 3].

Light-induced degradation, also called photodegradation, of wood occurs in its surface layer. An earlier study that employed electron spin resonance (ESR) estimated that UV and visible light penetrate wood to a depth of 75 and 200 µm, respectively [4]. Later works by Kataoka et al. reported that, on the basis of Fourier transform infrared (FT-IR) microscopy measurements, that the depth of light penetration into wood and the spreading rate of the degradation layer depend on the wavelength of light source [5], exposure time [6, 7], and wood density [8]. In addition, several other tools have also been employed to investigate the depth analysis of photodegraded wood [9,10,11,12]. Recently, high spatial resolution IR microscopy techniques such as infrared scanning near-field optical microscopy (IR-SNOM) and atomic force microscopy-based infrared spectroscopy (AFM-IR) have been developed, but there are only a few applications of these techniques to wood samples [13, 14].

Confocal Raman microscopy is useful for evaluating the molecular structure of various materials with a high spatial resolution [15]. In recent years, this technique has come to be used in topochemical studies of native [16, 17] and modified wood [18, 19]. Although some Raman methods such as FT-Raman [20] and UV resonance Raman spectroscopy [21] have been applied for monitoring of wood photodegradation, micro-scale investigation using the confocal system has not yet been performed. Therefore, in the present work, confocal Raman microscopy, which is helpful in investigating the micro-distribution of functional groups in wood cell walls, has been applied to the depth profiling analysis of light-exposed wood surfaces.

Materials and methods

Wood blocks measuring 140 (longitudinal) × 25 (radial) × 9 (tangential) mm were cut from the sapwood of air-dried sugi (Cryptomeria japonica D. Don). The radial surface of each block was smoothed with a wood planer and then exposed to artificial sunlight from a xenon arc light source (0.51 W/m² at 340 nm) in a weather-o-meter (Ci4000, Atlas, USA) with a black panel temperature of 65 °C and chamber temperature of 38 °C. After a 500-h exposure (no rain), 15-µm-thick cross sections used for Raman measurement (Fig. 1) were prepared on a sliding microtome.

Fig. 1

Procedure to obtain thin cross section along the light-exposed sample and the enlarged view of measurement area for depth profile analysis. L longitudinal, R radial, T tangential, asterisk excitation laser irradiated position

The depth profiling analysis was conducted using a confocal microRaman system (LabRAM ARAMIS, Horiba Jobin Yvon, France) equipped with a microscope (BX41, Olympus, Japan), a 100× objective lens (UPLSAPO, NA = 1.4, Olympus), and a diode-pumped solid-state laser (Ventus VIS 532, λ = 532 nm, Laser Quantum, UK). The incident laser power and the laser spot size on the sample were approx. 13 mW and 0.5 µm, respectively. Scattered Raman light was detected by a charge-coupled device (CCD) camera placed behind a 600-lines/mm grating. The Raman spectra were recorded in ten cycles with each cycle containing a 1 s integration time for one spot. Ten spectra were obtained and averaged, and the averaged spectra from ten different locations were again averaged. All the measurement positions were located on latewood, as shown in Fig. 1.

For the data acquisition and analysis, LabSpec 5 software (Horiba Jobin Yvon) was used. To remove the background from the fluorescence, the raw spectral data were baseline-corrected. Raman spectra were normalized by the intensity of the band at 1096 cm⁻¹ due to C–O and C–C stretching in polysaccharides [22].

Results and discussion

The Raman depth profiling spectra obtained from unexposed and exposed latewood are shown in Fig. 2. Significant changes in the Raman spectra were observed close to the surface. An obvious reduction in the intensities of the lignin bands at 1597 and 1139 cm⁻¹ assigned to aromatic ring stretching and coniferyl aldehyde structure [23], respectively, is observed. This was accompanied by an increase in the intensity of the carbonyl stretching [24] band at 1745 cm⁻¹. These spectral changes imply that degradation of lignin structures and formation of new carbonyl groups occurred in the wood surface during 500 h of exposure.

Fig. 2

Raman spectra for sugi latewood before (unexposed control) and after 500 h of exposure to artificial sunlight at different depths

To determine the maximum depth of photodegradation in latewood, the intensity ratios of the Raman bands at 1139, 1597 and 1745 cm⁻¹ against the reference band at 1096 cm⁻¹ were calculated and these are shown in Fig. 3. The filled symbols refer to statistically significant changes in intensity, and provide a measure of the maximum depth of photodegradation. Significant changes in the aromatic (1597 cm⁻¹), carbonyl (1745 cm⁻¹) and coniferyl aldehyde (1139 cm⁻¹) band intensity can be detected in latewood at the depths of 100–145, 249–285 and 426–496 µm, respectively.

Fig. 3

Changes in Raman intensities at a 1597 cm⁻¹, b 1745 cm⁻¹ and c 1139 cm⁻¹ as a function of depth for sugi latewood exposed to artificial sunlight. Filled symbols indicate that significant changes in peaks occurred as a result of light exposure (n = 20, difference between unexposed controls and exposed samples assessed determined by Smirnov–Grubbs test with a significance level of 0.05)

The degradation depth assessed by monitoring the carbonyl band (Fig. 3b) was somewhat greater than that observed in previous experiments using FT-IR microscopy [8]. This is due, in part, to the differences in measuring the position and exposed conditions because the depth of light penetration into wood and the spreading rate of degradation layer depends on the density of wood and the intensity and light exposure time [6,7,8].

The difference between the measured values obtained by FT-IR and Raman microscopy may also be related to the scale range of the instruments. The spatial resolution of Raman microscopy was higher than that of FT-IR microscopy. Raman depth profiling spectra in this study were recorded at each interval several tens of micrometers with a spot diameter of approx. 0.5 µm, while FT-IR depth profiling spectra in previous study were recorded at intervals of approx. 50 µm with a focusing area of 50 (depth) × 200 (width) µm. The boundary between degraded and undegraded areas can be separated with submicron precision by Raman microscopy leading to more accurate monitoring of photodegradation reactions that occur in wood surfaces.

It is important to highlight that the degradation depth estimated by the band of coniferyl aldehyde was extremely deep (Fig. 3c). A previous study that used resonance Raman spectroscopy revealed that coniferyl aldehyde is a photosensitive chromophore and as such, the structure of lignin is rapidly decomposed by light exposure [25]. In addition, blue light penetrates wood more deeply than UV light and is capable of bleaching wood, without causing marked changes in the IR spectra [5]. It is possible that the blue and longer wavelength light-induced partial degradation of lignin structures, such as the coniferyl aldehyde unit, which is undetectable by FT-IR, was reflected on the Raman spectra and, therefore, the degradation layer developed extremely deep into the region and could be detected. In a follow up report, we will investigate the effects of the wavelength of exposed light on degradation of wood constituents by Raman method.