3 RESULTS AND DISCUSSION
3.3 SET UP FOR DIVERSE GAS ANALYSIS
The Tenax® samples (200 mg) were desorbed (UltrA/ Unity, Markes Inc.) and detections were carried out by mass spectrometry, using GC/ MS, 7890A GC and 5975C MS (Agilent Technologies) and library spectra. Low aldehydes were identified by high-‐performance liquid chromatography (HPLC; EN ISO 16000-‐3). The DNPH cartridges were extracted with acetonitrile and analysed via HPLC using a variable wavelength detector (HPLC/ 1100 System, Agilent Technologies). The aldehyde content was automatically calculated from the obtained peak area. Low acids separations were performed with an ion chromatography system with a conductivity detector with ion suppression using an ion exclusion chromatography column (Strömberg 2013). The measured concentration of volatile compounds (Eq. 2) in the outlet air [µg/ m³] was converted (Eq. 3) into the area specific emission rate [µg/ m² * h].
Concentration [µg/m3] = Peak area [ae] / (RF [ae/ng] x sample vol. [L]) / RRF (Eq. 2) Area specific emission rate [µg/m2h] = Concentration [µg/m3] x 0.339 (Eq. 3)
RF = Toluene equivalent (value of control sample from internal calibration setup of SP) RRF = Relative Response Factor (relative to toluene) expresses rate in compound specific amount Area specific air flow rate (0.339)
4 RESULTS AND DISCUSSION
The emissions obtained from the micro-‐chamber measurements were described separately for each species on the first and the third day (Fig.3). Low aldehydes, i.e.
formaldehyde (FA) and acetaldehyde (AA) were detected in the untreated specimens of both species. Concerning the untreated, green samples of both species, hydrolysed acids did not emit in greater amounts, due to the moisture content (>70%) of the untreated samples. In addition the formation of formaldehyde was not catalysed through an acetic environment. The content of terpenes and aldehydes, and thus the sum of VOCs (sum of aldehydes > C2 and terpenes) for the impregnated and thermally modified samples, was far greater than that of the untreated samples. However, these compounds possess very little corrosivity towards materials like textiles, paper and metals.
Figure 3: Means of area specific emission rates of specimens (Fir and Alder, n = 2 /batch)
Contrary to expectation, identified compounds were refound after the thermal modification and drying process. However, all variants of Fir samples did not show any low acids, i.e. acetic and formic acid, neither at the first nor at the third day.
Unfortunately the results obtained from the Fir samples were not likely to give information about the influence of the different modification parameters on the chemical compounds. In contrast to the Fir samples, in the impregnated, thermally modified and kiln-‐dried specimens of Alder no FA or AA values were detected. Due to aldehydes (> C2), low amounts of VOC were detected in all thermally treated and dried variants of Alder samples. As far as the detection limits for substances with high contamination potential for individual display case construction materials were concerned, the low amount of acidity and yet a missed formation of formaldehyde in the Alder samples were very positive results.
5 CONCLUSIONS
The study considered direct emissions as well as secondary products, concerning the formation of acids and aldehydes, caused by thermo-‐hydrolysis. It was found that the FLEC is an appropriate instrumentation for the performance of the investigation of small-‐
sized samples, with a respectively low emission concentration. Similarly, the three different methods of analysis (GC/MS, IC and HPLC) were essential for the investigations.
Reasonable adjustments were achieved in applying the appropriate method. With this advanced modification process, especially the suppressed formation of formaldehyde and the minimized amount of acids, the applicability of thermally modified timber products in sensitive environments, e.g. health care institutions or museums, could be achieved.
6 REFERENCES
Birkeland M.J., Lorenz L., Wescott J.M., Frihart C.R. 2010: Determination of native (wood derived) formaldehyde by the desiccator method in particelboards generated during panel production. Holzforschung, Vol. 64,pp. 429-‐433, Walter de Gruyter, Berlin, New York.
Carlson F.E., Phillips E.K., Tenhaeff S.C., Detlefsen W.D. 1995: Study of formaldehyde and other organic emissions from pressing of laboratory oriented strandboard. Forest Products Journal J.45, 71-‐77.
Englund F. 2010: Neutral materials in the museum environment: Emissions from materials, SP Technical Research Institute of Sweden, Stockholm.
Roffael E., Hameed M., Kraft R. 2007: Bildung von Formaldehyd, Furfural und Ameisensäure bei der thermohydrolytischen Behandlung von einigen monomeren Zuckern (Xylose, Arabinose und Ga-‐lactose), Beitrag zu Entstehung von flüchtigen organischen Verbindungen (VOC) beim Holzaufschluss für die MDF-‐Herstellung, Holztechnologie, Bd. 48, Nr. 2, S. 15-‐18.
Schäfer M., Roffael E. 2000: On the formaldehyde release of wood. Holz als Roh-‐ und Werkstoff 58, S. 321-‐322.
Stamm A.J. 1964: Wood and cellulose science. Ronald Press, New York.
Strömberg N. 2013: SP Method for carboxylic acids in air, SP Chemistry and Materials Technology, Borås, Sweden.
ASTM D 5582-‐00 2006: Bestimmung der Formaldehydkonzentrationen aus Holzprodukten mit einem Exsikkator, Ausgabedatum: 2000, reapproved: 2006, Beuth, Berlin.
CARB 2008: California Air Resources Board, California Environmental Protection Agency.
State of California, USA.
DIN EN 717-‐1 2004: Holzwerkstoffe -‐ Bestimmung der Formaldehydabgabe -‐ Teil 1:
Formaldehydabgabe nach der Prüfkammer-‐Methode, Beuth, Berlin.
DIN EN 717-‐2 1994: Holzwerkstoffe -‐ Bestimmung der Formaldehydabgabe -‐ Teil 2:
Formaldehydabgabe nach der Gasanalyse-‐Methode, Beuth, Berlin.
EN 120 1993: Holzwerkstoffe -‐ Bestimmung des Formaldehydgehaltes -‐
Extraktionsverfahren (genannt Perforatormethode), Beuth, Berlin.
ISO 16000-‐10 2006: Indoor air -‐-‐ Part 10: Determination of the emission of volatile organic compounds from building products and furnishing -‐ Emission test cell method, SIS Förlag AB, 118 80 Stockholm.
Application of FT-NIR for recognition of substances used for conservation of wooden parquets of 19th century manor
houses located in South-Eastern Poland
Anna Rozanska1, Anna Sandak2
1 PhD student, Warsaw University of Life Sciences (SGGW), Nowoursynowska 166, 02-‐787 Warsaw, Poland, annamaria.rozanska@gmail.com
2 IVALSA Tree and Timber Institute, via Biasi 75, 38010 San Michele all’Adige (TN), Italy, anna.sandak@ivalsa.cnr.it
ABSTRACT
In antique palaces and manor houses, traditional surface finishing techniques were used.
The surfaces of antique wooden parquets were soaked with wax or with oils. Those substances preserve wood and have a major influence on its durability. FT-‐NIR spectral analysis of the surface of wood samples obtained from 19th century manor houses are one of the aspects of the ongoing research on antique parquet surface properties. The goal of this study was to verify if FT-‐NIR is useful technique for recognition of natural substances used for the wooden floor conservation.
Spectroscopy highlighted differences between different natural surfaces finishing methods. Even if algorithm works well for classification of contemporary wood, evaluation of antique floor was problematic and is under detailed investigation.
Keywords: FT-‐NIR spectroscopy, surface finishing: waxing and varnishing, surface quality, antique wooden parquets
1 SCIENTIFIC BACKGRAUND
The floor is an architectural element that, together with the walls and the ceiling, constitutes an integral part of the interior. It is as a valuable element of interior decoration. Unfortunately, due to the introduction of collective property and the
appropriation of manor houses after World War II, the parquets were irreversibly destroyed together with other elements of interior furnishing. There is an urgent need to carry out assessment and to document the buildings that have been preserved. Moreover such work might be useful to recreate the designs and structure of non-‐existing parquets and in consequence to preserve part of national heritage. It is therefore necessary to develop the knowledge related to their chemical, physical and mechanical properties, which will become the basis for their conservation programme.
2 SURFACE FINISHING
In antique palaces and manor houses, traditional surface finishing techniques were used.
The parquet had to be placed evenly and well levelled. After the wooden parquets were installed, they were scraped with a hand scraper along the fibres. Before manual smoothing, subsequent parts of the parquet were wetted with hot water and after smoothing were polished with steel shavings. Consequently the surface was soaked with wax or with oils. The parquet finishes has a major influence on its durability, since finishing techniques improves its resistance to abrasion (Rozanska et al. 2012a; 2012b;
2013).
3 PROJECT DESCRIPTION
3.1 FOURIER TRANSFORM NEAR INFRARED SPECTROSCOPY
It has been found that the energy of infrared is exciting particular parts of molecules in the surface of the matter and part of the energy are also absorbed. Different molecule combinations (such as C-‐H, O-‐H or N-‐H) are stimulated to vibrations depending on the molecular structure, chemical composition or physical properties of the surface measured (Coates 2000, Pasquini 2003, Tsuchikawa 2007). As an effect of this phenomenon the infrared radiation reflected from the surface can be used for estimation of the physical/chemical structure of the surface what has been a base for the measurements performed in this project. Advantage of FT-‐NIR is non-‐destructive character and relatively fast measurement, while the complex process of result interpretation is often
problematic. For this reason, the experiences of the IVALSA/CNR research centre (Trees and Timber Institute/National Research Council of Italy) were necessary to evaluate data.
3.2 GOAL OF THE PROJECT
FT-‐NIR spectroscopy was used for recognition of traditional finishing substances applied on antique wooden parquets of 19th century manor houses located in South-‐Eastern Poland. The other task was to create spectral database of substances commonly applied in 19th century. Finally verification of effectiveness of FT-‐NIR as a tool supporting conservation decisions was evaluated.
3.3 PROJECT WORKING PLAN
The project was carried out by two institutions (WULS-‐SGGW and IVALSA/CNR). Research tasks were divided into five working packages: 1. Preparation of samples acquired from antique wooden parquets and reference samples made of contemporary wood; 2.
Soaking contemporary wood samples with wax and varnish; 3. FT-‐NIR measureemnt of contemporary and antique samples; 4. Analysis of results by means of chemometric methods (development of models for contemporary wood and their verification; 5.
Assessment of chemical substances used for surface finishing.
3.4 CONTEMPORARY WOOD SAMPLES PREPARATION
Samples of different wood species: oak -‐Quercus sp, elm –Ulmus sp, ash-‐ Fraximus excelsior L and pine-‐ Pinus sylvestris L. were utilized for specimens preparation. To receive more objective results, contemporary wood from the same region of Poland as the antique wood was selected. Additionally wood was selected taking into account similar growth ring width, type of anatomical section and wood density.
The wood was cut into samples with dimensions of about 100x100x15mm. Before applying the coatings, the surface of samples was prepared by polishing with sand paper with grit of ca. 50-‐ 100-‐ 150. This task was performed manually, to avoid the changes to
the properties of wood surface that occur as a result of high temperature that is created during mechanical processing (Sandak et al. 2009).
Such prepared wood was soaked with varnish and wax. Natural bee wax was obtained from a bee yard in honeycombs. Wax was applied by pressing wax bars (made of melted bee wax) against the surface and then it was rubbed into it by polishing with a piece of felt. Linseed varnish (containing of 98% linseed oil and 2% siccatives) was prepared according to traditional recipe (Kinney 1971, Frid 1981). Varnish was applied hot, with the help of a brush, until the surface was entirely saturated. Samples prepared in that way served as reference data for determining the chemical substances applied for the antique parquet. Depth of finishing layer penetration also was investigated, since hot wax and varnish applied are not film-‐forming substances. All the measurement were carried out six month after the application of finishes to assure hardening of varnish. Before the tests, samples were acclimatised in standard climate conditions (± 20°C, ± 60% of relative air humidity).
3.5 ANTIQUE WOOD SAMPLES PREPARATION
The antique parquets come from buildings located in South-‐Eastern Poland. They date to the beginning of the 19th century – in case of the manor houses in Tarnowiec – and to the second half of the 19th century – in case of the manor house in Falejówka. All the parquets have been preserved on site in their original state and did not undergo comprehensive maintenance in the past. They are made of various wood species: oak or oak in combination with elm in case of 3 parquets from the Tarnowiec manor house and oak in case of a panel parquet from the Falejówka manor house. The antique parquet samples were taken from floors with different kinds of structure. In each room, samples were taken from three points: external corner of the room, traffic path and internal corner of the room. Ten samples were taken from each parquet floor (Table 1). Surface of the samples were refreshed (handy sanding) in order to remove layer of dirt and assure flatness of the samples.
Table 1: Characteristics of samples