Part of an indestructible component of any orthodox church, the Imperial Gates represent an important symbol in our cultural heritage. But in many cases the Imperial Gates from the wooden churches were damaged. In order to preserve and restore them, the scientific investigations of the Imperial Gate belonging to Nicula Monastery wooden church were performed by employing nondestructive and destructive methods. The wood essence was established, with its “health” status being investigated by FTIR (Fourier Transform Infrared) spectroscopy and DSC (Differential Scanning Calorimetry) thermal analysis. The painting materials employed by popular artists were determined by FTIR and XRF (X-ray fluorescence) spectroscopy as gypsum, calcite (rear background), lead white (Archangel Clothes), lead-minium (Archangel Clothes, leaf), iron oxide (Imperial Gate frame), malachite (green), Prussian blue (blue), orpiment (yellow), aliphatic, ester, and protein (probably egg yolk degradation products). Using similar colors as in the original artwork (resulting from the scientific investigation of the pigments) a 3D reconstruction has been performed. The restored Imperial Gates are placed in the old Nicula wooden church, being included into a tourist and religious circuit.
Multidisciplinary investigations of church’s wooden ceiling [
The aim of this paper was to investigate the painting materials employed for the Imperial Gates of the Nicula wooden church by several methods such as X-ray fluorescence and FTIR spectroscopy, DSC, and 3D scanning techniques [
The Imperial Gates (Figures
(a) Imperial Gates of the wooden church and some sampling points. (b) Imperial Gates of the wooden church (rear side) and two sampling points.
Small quantities of the painting materials, having various colors, and wood samples have been collected by an authorized conservation and restoration specialist in order to be analyzed using FTIR spectroscopy. The samples (both painting materials and wood samples) were ground in an agate mortar till an extra fine texture is obtained and then included into KBr pellets (KBr pure spectral powder, purchased from Sigma-Aldrich). FTIR spectra have been registered with a resolution of 4 cm−1 using a JASCO 6100 FTIR spectrometer (with Globar source and TGS detector at room temperature) in the 4000 to 400 cm−1 spectral domain by employing KBr pellet technique. The spectra were processed by
The white labels and the corresponding number for each painting material sample are presented in Figures
Differential Scanning Calorimetry (DSC) was carried out by means of a Shimadzu DSC-60 calorimeter, the sample (~2 mg) being heated in the range of 20–550°C with a heating rate of 10°C/min in crimped aluminum sample cell. The analysis was performed in static air atmosphere.
In the case presented in this paper for digitization we used two types of equipment: a 3D laser scanner which is able to acquire both 3D shape and texture of Imperial Gates and a photo camera with 20-megapixel sensor for image acquisition. The results obtained after data acquisition stage are 3D model, texture UV Map format, and high resolution photo resulting from combining multiple photos; in our case the image is 3800 × 12000 pixels.
The results of the XRF investigations are synthesized in Supplementary Table
Based on XRF analysis results one can propose the composition of the employed painting materials as lead-minium (Pb3O4), AsS (realgar), and Fe2O3 for red; malachite for green; cobaltite (CoAsS) or smaltite (CoAs2) for blue; lead carbonate (PbCO3) for white; and As2S3 (orpiment) for yellow.
FTIR spectra of different painting materials are presented in Figures
FTIR spectrum of the Imperial Gate (IG) rear background.
FTIR spectrum of red border (solid line) and of blue (dashed line) and of yellow (dot line) painting materials.
FTIR spectrum of two samples collected from the Imperial Gates: solid line, IG_10; dashed line, IG_11.
FTIR spectra of green two painting materials: solid line, IG_5 sample; dashed line, malachite; dash-dot line, IG_6 sample.
For IG_11 sample rear background the composition is calcium carbonate (1443 and 875 cm−1) as major component and gypsum (3442, 1154, and 596 cm−1) as minor component.
For red border, the painting materials composition is lead-minium and gypsum.
As compared to XRF data, one can propose smaltite, (Co,Fe,Ni)As2, (or cobaltite-CoAsS) for blue, taking into account the good collaboration of the local artists with the pigments’ merchants from Bohemia. One can eliminate azurite by comparing our FTIR spectrum with azurite one from RRUFF databank.
Sample 10 contains probably more PbCO3, then CaCO3 (and barium sulfate), and gypsum, whereas sample 11 contains more CaCO3 (and barium sulfate), then PbCO3, and gypsum. Samples 10 and 11 contain quartz traces, with a shoulder at ~1033 cm−1 being visible, also.
Besides gypsum, present in all painting materials, malachite could be a good component for green painting material. XRF data showed that Cu is major element; we proposed then malachite as green pigment; see Figure
FTIR spectra of wooden samples collected from the Imperial Gates compared to standard lime species are presented in Figure
FTIR spectra of Imperial Gates wood species as compared to lime species: solid line, historical up wooden stand; dashed line, Imperial Gates wood; dash-dot line, lime wood standard.
One can sustain that lime is the wood species employed for these Imperial Gates taking into account the similarities of the FTIR spectra, especially in the (3000–2700) and (2000–400) cm−1 spectral domains.
For analyzing the wood conservation state of the Nicula Imperial Gates, some parameters were determined as follows: the crystallinity indexes, defined [
Conservation state of lime wooden samples.
Sample | | | TCI | LOI | (L/C)1 | (L/C)2 | (L/C)3 | (L/C)4 |
---|---|---|---|---|---|---|---|---|
Historical wooden stand | 1.07 | 0.98 | 1.26 | 0.59 | 1.19 | 0.57 | 1.13 | 0.76 |
Historical up wooden stand | 1.02 | 0.71 | 1.19 | 0.43 | 0.93 | 0.56 | 1.05 | 0.75 |
Modern lime wood | 1.21 | 3.11 | 1.25 | 1.71 | 0.71 | 0.44 | 1.15 | 0.61 |
The crystallinity is decreased for historical lime wood as compared to modern one; see the corresponding values from Table
Two exotherms were observed in the DSC curves of standard lime wood at 330 and 440°C (see Figure
DSC curves of Nicula Imperial Gates and of lime species standard wood.
The maximum temperature of the characteristic peaks increases with ~15°C [
The digital restoration of the Imperial Gates is done in four steps. Digitization of Imperial Gates: in this phase, we use different techniques to obtain the 3D model of Imperial Gate. Digital restoration of wood support: using CAD software (Catia V5), the wood support is restored and completed with the missing parts. Digital restoration of painted layer I: using photo editing software, the painted layer is restored to the actual stage. Digital restoration of painted layer II: using results from XRF and FTIR, the painted layer is restored with the original colors.
After digital restoration process is completed, the two components are assembled into a 3D model that can be used in web dissemination of the cultural heritage assets.
In first stage digitization of the Imperial Gates was done using two different methods: laser scanning and photogrammetry. We obtain two different 3D models (see Supplementary Figure
The 3D models obtained using the two methods of 3D digitization are in the form of a triangular mesh (see Figure
The restoration of the painted layer with the current colors.
In the restoration of the painted layer we can have two distinct situations: the restoration with the current colors (see Figure
Having the painted layer digitally restored and the wooden support 3D model fully assembled with all the missing elements, the final composition for the whole digitally reconstructed 3D model can be assembled. Depending on the final destination and purpose of the 3D model, we can have two situations.
The first one involves the creation of a high definition 3D digital model using the model obtained using the laser scanning equipment and using the painted layer. This 3D model can be used in the 3D digital documentation and further analysis of the restoration process since its geometry is highly accurate and it represents the real object with very high fidelity.
The second situation involves the creation of a low definition 3D model that uses the mesh obtained from the photogrammetry 3D digitization process and it is textured with the digitally restored paint layer. This model is best suited for online dissemination since the file size will be optimized for in browser visualization and the required computer resources needed for this process are not too high allowing a higher number of potential users to have access to this 3D model.
Taking into account all these investigations, the Imperial Gates were preserved and restored by specialists and the result of these processes is presented in Figure
Restored Imperial Gate (left side) of Nicula wooden church Imperial Gates.
Based on the XRF and FTIR spectroscopy the painting materials employed as painting materials for Imperial Gates are lead-minium, AsS (realgar) and Fe2O3 for red; cobaltite or smaltite for blue; malachite for green; lead carbonate for white; As2S3 for yellow; and PbCO3, CaCO3, and gypsum as background and for white color. FTIR and DSC measurements conclude that lime is the wood species employed for Imperial Gates. Its conservation status, taking into account the increase of the amorphous content, is established by FTIR analyses that sustained a decrease of the lime wood crystallinity during time; increased consumption of cellulose was also established. Digital restoration of Imperial Gates can be a solution for digital preservation of this part of our cultural heritage. Using low definition 3D model the dissemination of Imperial Gates can be done using web distribution and also mobile compatible format. An example of web distribution can be seen on the project website:
I. Bratu and C. Măruţoiu are co-first authors.
The authors declare that there is no conflict of interests regarding the publication of this paper.
The financial support of PN II-PT-PCCA-2013-4-1882 project is greatly acknowledged.