Impact of Animal-Based Glues on the Surface Characteristics of Traditional Wood-Supported Polychrome Coatings

Abstract

Animal-derived adhesives have historically played a critical role in East Asian polychrome artworks, yet their influence on coating properties remains insufficiently understood. In this study, three traditional animal glues—bone glue, hide glue, and fish glue—were systematically compared as binding media for cinnabar (red), orpiment (yellow), and lazurite (blue) coatings on pinewood substrates. FTIR spectra of all glues exhibited characteristic collagen-related absorption bands, with bone glue showing a stronger hydroxyl peak at 3280 cm⁻¹. Viscosity analysis revealed bone glue as the most stable, reaching 120 ± 6 mPa·s at 20% concentration, compared with lower values for fish glue and hide glue. Colorimetric analysis indicated that cinnabar coatings were primarily affected by glue concentration, while orpiment coatings showed significant glue-type effects on chroma, with hide glue producing the highest saturation. Lazurite coatings displayed the strongest adhesive-related differences: hide glue enhanced saturation, whereas fish glue increased lightness. Gloss remained consistently low (<3 Gloss Units (GUs)), in line with the matte appearance of historical coatings, although bone glue produced smoother cinnabar surfaces (Ra = 1.72 µm, GU = 3.09). Two-way ANOVA confirmed that both the glue type and concentration significantly influenced chroma, gloss, and roughness. These findings demonstrate that although all animal glues share a collagenous origin, their distinct physicochemical properties shape the optical and microstructural qualities of polychrome coatings, offering a scientific basis for adhesive selection in cultural heritage conservation.

1. Introduction

Polychrome coatings, consisting of pigments bound by organic adhesives and applied to wooden substrates, represent one of the most distinctive artistic and cultural expressions in East Asia. These materials are not only of aesthetic significance but also serve as important records of historical craftsmanship and technological development. Among the various binders historically employed, animal-derived glues—particularly bone glue, hide glue, and fish glue (fish bladder glue)—have been widely used due to their strong adhesion, workability, and availability [1,2]. Derived from collagen, these glues possess structural features that enable them to act as effective binding media, but they are also susceptible to environmental fluctuations and biological degradation [3,4].

In recent years, advances in heritage science have significantly improved our understanding of proteinaceous adhesives in artworks. Studies using proteomics, FTIR, and chromatographic approaches have confirmed that collagen-based glues dominate historical coatings and that their chemical signatures can vary depending on animal origin and processing conditions [5,6]. Furthermore, rheological properties such as viscosity directly influence pigment dispersion and coating stability, which ultimately affect visual qualities including color, gloss, and surface morphology [7,8].

Nevertheless, previous studies such as Schellmann (2013) [9] and Pellegrini et al. (2016) [10] have already explored certain aspects of animal glue identification and behavior in coatings. What distinguishes the present work is its systematic evaluation of three glue types at multiple concentrations, applied to traditional pigments on wooden supports, with combined optical and surface characterization. This approach enables a more quantitative understanding of how adhesive type and dosage interactively influence chromatic and microstructural properties, going beyond prior descriptive or single-parameter analyses.

The choice of pinewood as the experimental substrate is also significant. Pine (Pinus spp.) has historically been one of the most common supports in painted objects and architectural decorations, but its high content of resins and extractives makes adhesion particularly challenging [11,12]. Including pinewood substrates therefore allows this study to better replicate the practical issues faced in historical polychrome systems.

Equally important is the selection of cinnabar (HgS), orpiment (As₂S₃), and lazurite (Na₈–₁₀Al₆Si₆O₂₄S₂–₄) as representative pigments. These materials were among the most prestigious and widely used in East Asian polychromy: cinnabar for its brilliant red hue, orpiment for its luminous yellow, and lazurite for its deep blue derived from lapis lazuli. Their inclusion ensures cultural relevance while also providing contrasting optical properties, making them ideal test pigments for adhesive-pigment interactions [13,14].

However, several challenges remain unresolved. First, comparative studies of different animal glues (bone, hide, and fish) under controlled experimental conditions are limited, particularly regarding their effects on chromatic parameters and surface characteristics of polychrome coatings. Second, while many studies have focused on chemical identification, relatively fewer have quantitatively examined how adhesive type and concentration jointly influence coating performance [15,16]. Lastly, the conservation community continues to debate the optimal adhesive choice for restoration interventions, balancing authenticity, durability, and reversibility [17].

The present study aims to address these gaps by systematically investigating the effects of bone glue, hide glue, and fish glue on the physicochemical and optical properties of polychrome coatings. Specifically, we combine FTIR analysis, viscosity testing, colorimetry, gloss measurement, and surface roughness evaluation to clarify how glue type and concentration affect coating performance. By providing comparative experimental data, this work seeks to inform conservation practice with a scientific basis for selecting historically appropriate adhesives, thereby contributing to both cultural heritage preservation and materials science.

2. Materials and Methods

2.1. Materials

Three traditional animal-based adhesives, namely bone glue (granular form), hide glue (granular form), and fish glue (flake form, fish bladder glue), each weighing 100 g, were selected for this study and supplied by Fuchuan Yao Autonomous County Caiyuan Wood Processing Factory, Hezhou, China). Three natural mineral pigments were employed: lazurite (blue, 100 g), orpiment (As₂S₃, yellow, 100 g), and cinnabar (HgS, red, 100 g), all purchased from Yongzhou Shizhicolor Industrial Pigment Co., Ltd., Yongzhou, Hunan, China, with a purity of no less than 98% and stored in sealed, light-protected containers under dry conditions. The wooden substrates consisted of pinewood plates (Pinus koraiensis, 100 mm × 100 mm × 8 mm, 10% ± 2% moisture content) without visible defects, also supplied by Fuchuan Yao Autonomous County Caiyuan Wood Processing Factory. The choice of animal glues and mineral pigments was guided by their historical significance in East Asian polychrome artworks, ensuring both authenticity and relevance to cultural heritage conservation studies [18].

2.2. Preparation of Polychrome Coatings

The preparation of animal glue solutions involved weighing a specific amount of adhesive (bone glue, hide glue, or fish glue) and dissolving it in distilled water in a glass beaker. Each mixture was heated in a thermostatic water bath and continuously stirred until the adhesive was completely dissolved. After cooling to room temperature, a yellowish-brown viscous solution was obtained. Three different concentrations of each adhesive solution were prepared: 10%, 15%, and 20% (w/w).

For pigment preparation, lazurite (blue), orpiment (yellow), and cinnabar (red) powders were separately mixed with the adhesive solutions at a pigment-to-glue solution mass ratio of 1:1. Each mixture was manually stirred for 30 min until a homogeneous paste without visible aggregates was achieved.

Pinewood substrates were pretreated by sanding with 400-grit abrasive paper and cleaned with anhydrous ethanol, followed by conditioning in a constant temperature–humidity chamber (25 °C, 50% RH) until a constant weight was reached. The pigment–glue pastes were evenly applied onto the wood surface using a coating bar (Zehntner ZAA 2300, Zehntner Testing Instruments, Sissach, Switzerland), with a controlled thickness of 0.15 ± 0.02 mm verified by a digital micrometer (Mitutoyo 293 Series, Mitutoyo Corp., Kawasaki, Kanagawa, Japan). For each formulation (glue type × concentration × pigment), three replicate samples were prepared. The choice of three replicates was based on widely adopted coating evaluation standards, and preliminary tests confirmed stable values with this sample size. In total, 81 specimens were produced. All coated samples were air-dried under dark conditions at 25 ± 1 °C and 50% ± 5% RH for 72 h to obtain stable polychrome films suitable for further testing. Artificial accelerated aging was not included in this study, as this work forms part of a master’s thesis; follow-up research will address aging effects and their implications [19].

2.3. Characterization of Animal Glues

To evaluate the physicochemical properties of bone glue, hide glue, and fish glue, two analytical methods were employed. Fourier-transform infrared spectroscopy (FTIR, Nicolet iS10, Thermo Fisher Scientific, Waltham, MA, USA, KBr pellet method) was conducted to characterize the functional groups of the animal-derived adhesives in their solid states. Each dried glue sample was finely ground and mixed with spectroscopic-grade KBr at a ratio of 1:100 (w/w), then pressed into a transparent pellet for measurement. Spectra were collected in the range of 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹. Characteristic protein-related absorption bands, such as Amide I (∼1650 cm⁻¹, C=O stretching vibration), Amide II (∼1550 cm⁻¹, N–H bending and C–N stretching), and Amide III (∼1240 cm⁻¹, C–N stretching and N–H deformation), were analyzed to confirm the collagen-based composition of the adhesives.

In addition, the viscosity of the glue solutions with different concentrations (10%, 15%, and 20% w/w) was measured using a digital viscometer (NDJ-5S, Shanghai Precision Instrument Co., Ltd., Shanghai, China) at 25 °C. Each measurement was repeated three times to ensure reproducibility, and mean values were reported. FTIR and viscosity testing are standard approaches for assessing the structural integrity and rheological behavior of proteinaceous adhesives, which are critical for understanding their performance as binding media in traditional polychrome coatings [20,21,22,23].

2.4. Evaluation Methods

2.4.1. Appearance and Color Evaluation

To compare the effects of glue type and concentration on the surface appearance and color of the coatings, one representative specimen from each group was scanned using a Canon CanoScan LiDE 400 flatbed scanner (Canon Inc., Tokyo, Japan). From each scanned image, a 5 cm × 5 cm region was cropped for direct visual comparison.

The chromatic parameters of the polychrome coatings were determined using a colorimeter (HP-200, Shanghai Hanpu Co., Ltd., Shanghai, China) under standard illuminant D65 and a 10° observer angle. Measurements were taken at five random positions on each sample surface, and the average values were recorded in the CIE Lab* color space, including lightness (L*), red–green coordinate (a*), and yellow–blue coordinate (b*). The chroma (C* = (a*² + b*²)^1/2) was calculated to evaluate the color saturation of each coating. The overall color difference (ΔE) was further calculated relative to a white calibration plate according to the CIE76 formula [24].

2.4.2. Gloss Measurement

Surface gloss was evaluated using a glossmeter (NHG268, Shenzhen 3nh Technology Co., Ltd., Shenzhen, China) at an incident angle of 60°. Three measurements were performed on each coating, and mean values were reported. Gloss testing provides insight into surface reflectance and visual finish quality, which are important indicators of the aesthetic characteristics of traditional polychrome coatings [25].

2.4.3. Surface Roughness and Morphology

The surface roughness of the coatings, characterized by parameters Ra, Rq, and Rt, was quantified with a JB-4C roughness tester (Taiming Optical Instrument Co., Ltd., Shanghai, China). Measurements were taken at three randomly selected locations per specimen, and the results are reported as average values. This quantitative analysis of roughness aids in understanding how pigment morphology, glue type, and concentration collectively influence the surface microstructure [26].

Complementing this, pigment dispersion was qualitatively assessed through optical microscopy. Observations were carried out using a Zeiss Axio Scope A1 microscope (Carl Zeiss AG, Oberkochen, Germany) in reflected light mode at various magnifications (40× to 100×), with the integrated AxioVision Rel. 4.8 software employed for image acquisition.

2.5. Statistical Analysis

All experiments were conducted in triplicate, and results are expressed as mean ± standard deviation (SD). Statistical analyses were performed using SPSS 25.0 software (IBM, Armonk, NY, USA). A two-way analysis of variance (ANOVA) was applied to evaluate the significance of two independent factors—animal glue type (bone glue, hide glue, and fish glue) and glue concentration (10%, 15%, and 20%)—on the measured parameters, including chromatic values (L*, a*, b*, C*, and ΔE), gloss, and surface roughness indices (Ra, Rq, and Rt). Interaction effects between glue type and concentration were also examined. Post hoc comparisons were carried out using Tukey’s test, and a p-value < 0.05 was considered statistically significant [27].

3. Results

3.1. Physicochemical Properties of Animal Glues

The FTIR spectra of bone glue, hide glue, and fish glue are shown in Figure 1. All three adhesives exhibited similar characteristic absorption bands, confirming their proteinaceous and collagen-based composition. A broad and strong band was observed around 3280 cm⁻¹, attributed to O–H and N–H stretching vibrations, with bone glue displaying a more intense peak at this position compared with hide glue and fish glue. The band near 3070 cm⁻¹ corresponded to amide B (N–H stretching), while the peaks at 2935 and 2855 cm⁻¹ were associated with C–H stretching of aliphatic groups. The amide I band at approximately 1630 cm⁻¹ (C=O stretching), the amide II band at 1530 cm⁻¹ (N–H bending and C–N stretching), and the amide III band at 1270–1230 cm⁻¹ confirmed the collagen backbone of the animal glues. Additional absorption peaks at 1447 cm⁻¹ (CH₂ bending), 1326 cm⁻¹ (amide III contribution), 1080 cm⁻¹ (C–O stretching), and 726 cm⁻¹ (skeletal vibrations) were also present in all samples. Although the three adhesives shared a similar spectral profile, differences were noted in peak intensity: bone glue showed a stronger hydroxyl-related band at 3280 cm⁻¹, suggesting higher hydrogen bonding capacity. Fish glue showed slightly more defined amide I and II peaks, though these differences were subtle and should be interpreted cautiously. Hide glue presented intermediate spectral characteristics, which may reflect partial collagen hydrolysis occurring during its industrial extraction and purification processes [10,28,29]. These observations are in agreement with prior reviews emphasizing the sensitivity of amide band intensity to collagen conformation and source [30].

The viscosity of the three animal glues at different concentrations (10%, 15%, and 20%) is presented in Figure 2. All adhesives exhibited a positive correlation between concentration and viscosity, consistent with their protein-based gel-forming nature. Specifically, the viscosity values for bone glue were 72 ± 6, 78 ± 6, and 120 ± 6 mPa·s at 10%, 15%, and 20% concentrations, respectively; for hide glue, 66 ± 6, 78 ± 6, and 90 ± 6 mPa·s; and for fish glue, 72 ± 6, 84 ± 6, and 98 ± 9 mPa·s. Among the three adhesives, bone glue showed the highest viscosity increase at 20%, indicating stronger gelation capability and intermolecular interactions at high concentrations. Fish glue exhibited moderate viscosity values across concentrations, but no definitive conclusions about its molecular network can be drawn solely from viscosity analysis. Hide glue, in contrast, exhibited the lowest viscosity overall, which may be attributed to partial collagen hydrolysis and reduced molecular weight. It is also noted that the viscosity increase with concentration was approximately linear for hide and fish glues, whereas bone glue showed a more pronounced, nonlinear rise at 20% (Figure 2). These results demonstrate that although all animal glues share a collagen-based structure, their rheological behaviors differ significantly depending on molecular integrity and gelation potential [31].

3.2. Color Characteristics of Polychrome Coatings

Representative scanned images of the painted specimens are presented in Figure 3, providing a direct visual comparison of surface appearance for different pigments prepared with three glues at different concentrations. These images highlight the overall visual differences before quantitative evaluation.

The chromatic parameters of polychrome coatings prepared with bone glue, hide glue, and fish glue are summarized in

Table 1. Color parameters (mean ± SD) of red polychrome coatings prepared with different animal glues and concentrations.

Factor	L*	a*	b*	C*
Glue Type
Bone Glue	44.47 b ± 0.4	47.58 a ± 0.39	24.11 a ± 0.31	53.36 a ± 0.44
Hide Glue	43.34 a ± 0.3	46.42 a ± 0.36	23.99 a ± 0.33	52.26 a ± 0.41
Fish Glue	43.18 a ± 0.40	46.39 a ± 0.37	24.48 a ± 0.35	52.46 a ± 0.42
Concentration
10%	45.01 c ± 0.45	47.87 b ± 0.42	25.61 c ± 0.36	54.30 c ± 0.52
15%	44.22 b ± 0.41	46.68 a ± 0.39	24.04 b ± 0.35	52.51 b ± 0.47
20%	41.76 a ± 0.38	45.84 a ± 0.37	22.92 a ± 0.34	51.26 a ± 0.45
p Values
Glue Type	0.002	0.090	0.367	0.162
Concentration	<0.001	0.009	<0.001	<0.001
Glue Type × Concentration	<0.001	<0.001	0.003	<0.001

Note: Mean ± SD values (n = 45) of L, a, b and C followed by the same small superscript letters (a–c) within a group are not significantly different based on Fisher’s Protected LSD test at the 0.05 significance level.

,

Table 2. Color parameters (mean ± SD) of yellow polychrome coatings prepared with different animal glues and concentrations.

Factor	L*	a*	b*	C*
Glue Type
Bone Glue	78.99 a ± 0.52	12.97 b ± 0.29	67.30 a ± 0.48	68.54 a ± 0.50
Hide Glue	78.42 a ± 0.49	12.96 b ± 0.28	69.82 b ± 0.50	71.02 b ± 0.52
Fish Glue	79.09 a ± 0.50	12.22 a ± 0.27	69.19 b ± 0.49	70.27 b ± 0.51
Concentration
10%	81.60 c ± 0.63	11.54 a ± 0.26	68.10 a ± 0.47	69.08 a ± 0.49
15%	78.41 b ± 0.55	13.37 b ± 0.30	70.61 b ± 0.52	71.87 b ± 0.54
20%	76.49 a ± 0.54	13.23 b ± 0.29	67.60 a ± 0.46	68.89 a ± 0.48
p Values
Glue Type	0.270	<0.001	0.011	0.013
Concentration	<0.001	<0.001	0.002	0.001
Glue Type × Concentration	0.002	<0.001	<0.001	<0.001

Note: Mean ± SD values (n = 45) of L, a, b and C followed by the same small superscript letters (a–c) within a group are not significantly different based on Fisher’s Protected LSD test at the 0.05 significance level.

and

Table 3. Color parameters (mean ± SD) of blue polychrome coatings prepared with different animal glues and concentrations.

Factor	L*	a*	b*	C*
Glue Type
Bone Glue	46.33 a ± 0.39	−3.76 b ± 0.21	−16.02 b ± 0.34	16.47 a ± 0.34
Hide Glue	52.71 b ± 0.42	−4.27 ab ± 0.23	−18.01 a ± 0.36	18.53 b ± 0.37
Fish Glue	57.47 c ± 0.41	−4.49 a ± 0.22	−16.19 b ± 0.35	16.81 a ± 0.36
Concentration
10%	57.40 c ± 0.47	−4.14 ab ± 0.23	−18.10 a ± 0.37	18.60 c ± 0.38
15%	53.40 b ± 0.44	−4.67 a ± 0.24	−17.03 b ± 0.35	17.66 b ± 0.36
20%	45.17 a ± 0.43	−3.70 c ± 0.21	−15.09 c ± 0.33	15.55 a ± 0.35
p Values
Glue Type	<0.001	0.043	<0.001	<0.001
Concentration	<0.001	0.009	<0.001	<0.001
Glue Type × Concentration	0.007	0.027	<0.001	<0.001

Note: Mean ± SD values (n = 45) of L, a, b and C followed by the same small superscript letters (a–c) within a group are not significantly different based on Fisher’s Protected LSD test at the 0.05 significance level.

. Results demonstrated that both glue type and concentration exerted significant effects on color development, with distinct trends observed among red, yellow, and blue coatings.

The results for cinnabar-based coatings are p resented in Table 1. The lightness (L*) values showed significant differences among adhesives (p = 0.002), with bone glue producing slightly higher L* than hide glue and fish glue. However, a* and b* values, as well as chroma (C*), showed no significant differences between glue types, suggesting that glue type exerted only a minor influence on the red hue. In contrast, glue concentration had a pronounced effect (p < 0.001). At 10%, coatings appeared brighter and more saturated (L* = 45.01, C* = 54.30), whereas at 20%, the coatings became noticeably darker (L* = 41.76, C* = 51.26). These results indicate that pigment concentration and binder content interacted strongly in determining the vividness of red coatings.

As shown in Table 2, yellow coatings exhibited high lightness values across all formulations. glue type had limited influence on L* (p = 0.270) but significantly affected a*, b*, and C* values. Hide glue and fish glue tended to produce stronger yellowness (b* ≈ 69–70) and higher saturation compared with bone glue. Concentration effects were more pronounced: at 10%, coatings were the brightest (L* = 81.60) with moderate saturation, while at 20%, they became darker (L* = 76.49) yet showed higher a* values. These findings suggest that both adhesive type and concentration jointly influenced the chromatic stability of yellow coatings.

The most pronounced glue-type effect was observed in lazurite-based coatings (Table 3). Hide glue produced the darkest and most saturated blue (L* = 52.71, C* = 18.53), whereas bone glue resulted in lower lightness and saturation (L* = 46.33, C* = 16.47). Fish glue yielded the highest L* (57.47) but relatively low chroma (C* = 16.81), indicating a lighter yet less vivid tone. Concentration strongly influenced all parameters (p < 0.001). At 10%, coatings were the brightest (L* = 57.40, C* = 18.60), while at 20%, they appeared significantly darker (L* = 45.17, C* = 15.55). Interaction effects were also significant, suggesting that the relationship between glue type and pigment dispersion strongly affected blue color development [32].

3.3. Gloss and Surface Roughness Analysis of Polychrome Coatings

The gloss (GU) and surface roughness parameters (Ra, Rq, and Rt) of the polychrome coatings are summarized in

Table 4. Surface roughness and gloss of red polychrome coatings.

Factor	Ra	Rq	Rt	GU
Glue Type
Bone Glue	1.72 a ± 0.12	2.71 a ± 0.18	19.02 a ± 1.05	3.09 b ± 0.21
Hide Glue	3.50 c ± 0.25	4.99 b ± 0.35	28.74 b ± 1.85	0.87 a ± 0.08
Fish Glue	2.77 b ± 0.19	4.34 b ± 0.30	24.46 b ± 1.52	0.92 a ± 0.09
Concentration
10%	2.41 a ± 0.17	3.45 a ± 0.24	19.85 a ± 1.21	0.90 c ± 0.07
15%	2.48 a ± 0.18	3.68 a ± 0.26	25.13 b ± 1.65	1.31 b ± 0.11
20%	3.50 b ± 0.24	4.90 b ± 0.34	27.22 b ± 1.78	2.67 c ± 0.19
p Values
Glue Type	<0.001	<0.001	0.003	<0.001
Concentration	<0.001	0.017	0.017	<0.001
Glue Type × Concentration	<0.001	0.912	0.209	<0.001

Note: Mean ± SD values (n = 27) of Ra, Rq, Rt and GU followed by the same small superscript letters (a–c) within a group are not significantly different based on Fisher’s Protected LSD test at the 0.05 significance level.

,

Table 5. Surface roughness and gloss of yellow polychrome coatings.

Factor	Ra	Rq	Rt	GU
Glue Type
Bone Glue	3.57 b ± 0.21	5.27 a ± 0.29	26.90 a ± 1.45	2.16 b ± 0.14
Hide Glue	3.27 a ± 0.20	5.23 a ± 0.28	27.02 a ± 1.46	2.06 a ± 0.13
Fish Glue	3.56 b ± 0.21	5.70 a ± 0.31	29.44 a ± 1.58	1.79 b ± 0.12
Concentration
10%	3.06 a ± 0.18	4.61 a ± 0.25	22.98 a ± 1.25	2.02 b ± 0.13
15%	3.85 b ± 0.23	6.06 b ± 0.33	30.52 b ± 1.64	1.86 a ± 0.12
20%	3.58 c ± 0.21	5.53 a ± 0.30	29.86 b ± 1.60	2.12 b ± 0.14
p Values
Glue Type	0.020	0.572	0.356	<0.001
Concentration	<0.001	0.024	0.002	<0.001
Glue Type × Concentration	<0.001	0.002	0.002	0.013

Note: Mean ± SD values (n = 27) of Ra, Rq, Rt and GU followed by the same small superscript letters (a–c) within a group are not significantly different based on Fisher’s Protected LSD test at the 0.05 significance level.

and

Table 6. Surface roughness and gloss of blue polychrome coatings.

Factor	Ra	Rq	Rt	GU
Glue Type
Bone Glue	3.44 a ± 0.21	5.24 a ± 0.29	28.76 a ± 1.58	0.90 a ± 0.06
Hide Glue	4.17 a ± 0.25	6.19 a ± 0.34	32.24 a ± 1.77	1.00 a ± 0.07
Fish Glue	3.57 a ± 0.21	5.15 a ± 0.28	28.72 a ± 1.58	0.96 a ± 0.07
Concentration
10%	4.05 a ± 0.24	5.65 a ± 0.31	30.03 a ± 1.65	1.07 a ± 0.07
15%	3.68 a ± 0.22	5.71 a ± 0.31	29.06 a ± 1.60	0.94 b ± 0.06
20%	3.46 a ± 0.21	5.22 a ± 0.29	30.63 a ± 1.68	0.84 b ± 0.06
p Values
Glue Type	0.065	0.124	0.109	0.195
Concentration	0.179	0.607	0.686	0.002
Glue Type × Concentration	0.057	0.095	0.020	0.106

Note: Mean ± SD values (n = 27) of Ra, Rq, Rt and GU followed by the same small superscript letters (a, b) within a group are not significantly different based on Fisher’s Protected LSD test at the 0.05 significance level.

. Overall, all samples exhibited very low gloss values (< 3 GU), consistent with the matte appearance typical of traditional polychrome finishes. Surface roughness values varied depending on pigment type, glue type, and concentration. As shown in Table 4, significant differences were observed among animal glues (p < 0.001). Bone glue produced the smoothest surfaces (Ra = 1.72 µm, Rq = 2.71 µm) with the highest gloss (3.09 GU), whereas hide glue led to the roughest surfaces (Ra = 3.50 µm, Rq = 4.99 µm) and lowest gloss (0.87 GU). This conclusion is visually supported by optical micrographs (Figure 4a–i), which reveal a more continuous and uniform surface for bone glue coatings compared to the pronounced microtopographical features in hide glue samples. The images provide direct visual evidence of the microstructural differences correlating with surface roughness measurements. Fish glue showed intermediate values. Glue concentration had a significant effect on both roughness and gloss (p < 0.05). At 10% and 15%, coatings remained relatively smooth and matte, while at 20%, roughness increased substantially (Ra = 3.50 µm), and gloss also rose (2.67 GU). These results indicate that bone glue favored smoother and more reflective surfaces, while higher concentrations generally increased microtopographic irregularities, thus enhancing gloss variation.

The results for yellow coatings are given in Table 5. The glue type influenced Ra (p = 0.020) and gloss (p < 0.001) but had a negligible effect on Rq and Rt. Bone glue and fish glue resulted in higher Ra values (~3.56–3.57 µm) compared with hide glue (3.27 µm). Gloss was slightly higher for bone glue (2.16 GU) than for hide glue (2.06 GU) or fish glue (1.79 GU). Concentration effects were significant (p < 0.05), with 15% and 20% samples showing rougher surfaces (Ra = 3.85 and 3.58 µm, respectively) compared with 10% (3.06 µm). Gloss decreased slightly at 15% (1.86 GU) but increased again at 20% (2.12 GU). These findings suggest that hide glue favored smoother yellow coatings, while bone glue enhanced gloss.

As shown in Table 6, no significant differences in Ra, Rq, or Rt were observed among glue types (p > 0.05), although hide glue tended to produce slightly rougher surfaces (Ra = 4.17 µm). Gloss values were consistently low (< 1.1 GU) across all adhesives. Concentration significantly affected gloss (p = 0.002) but not roughness. At 10%, coatings had the highest gloss (1.07 GU), while at 20%, gloss dropped to 0.84 GU. These results indicate that pigment morphology was the dominant factor for blue coatings, while adhesive type and concentration only slightly modified surface reflectance.

4. Discussion

The present study systematically investigated the impact of three traditional animal-derived adhesives—bone glue, hide glue, and fish glue—on the chromatic, gloss, and microtopographic properties of polychrome coatings. FTIR analysis verified that all three glues were collagen-based, which is consistent with the established literature on animal glues [33,34]. Observed differences in band intensities, particularly the stronger 3280 cm⁻¹ band in bone glue, indicated subtle compositional variations. These findings align with previous studies showing that the collagen source (mammalian vs. fish) influences hydrogen bonding and protein conformation, thereby affecting adhesive performance in heritage applications [35].

Viscosity measurements further revealed that bone glue exhibited the highest viscosity at higher concentrations, followed by fish glue and hide glue. This order correlates with their structural stability and crosslinking density, properties that are known to determine adhesive film formation and pigment dispersion [4,28]. Such rheological differences likely contributed to the distinct chromatic responses observed among the three pigments. To conclusively determine whether the observed color differences originate from physical dispersion variations or potential chemical alterations, future work incorporating techniques with higher molecular specificity, such as Raman spectroscopy, would be highly beneficial.

Colorimetric analysis demonstrated that red cinnabar coatings were mainly affected by glue concentration rather than glue type, with higher concentrations resulting in darker and less saturated surfaces. In contrast, yellow orpiment coatings displayed significant glue-type effects, with hide glue enhancing chroma compared to bone glue. The most pronounced differences were found in blue lazurite coatings, where hide glue enhanced saturation, while fish glue increased lightness but decreased chroma. These outcomes highlight the complex interplay between pigment morphology and adhesive matrix, echoing recent findings that proteinaceous binders strongly mediate pigment-particle interactions and optical outcomes [32,36].

Gloss and surface roughness analyses confirmed that coatings generally retained a matte finish, characteristic of traditional polychrome artworks. The surface morphology was further elucidated by optical microscopy (Figure 4), which corroborated the roughness data. The analysis confirmed that glue type influenced surface microstructure: bone glue yielded smoother and more reflective red surfaces, while hide glue increased roughness, particularly in yellow coatings. Notably, higher glue concentrations consistently increased surface irregularities, which in turn affected gloss perception. These observations are aligned with recent conservation reports, which emphasize that protein binder concentration is a critical factor in balancing coating smoothness, durability, and authenticity in restoration practices [37].

It is important to acknowledge that this study focused on the initial properties of the coatings. The absence of artificial ageing limits the extrapolation of these findings to the long-term behavior of the polychromy. Future research incorporating ageing protocols is essential to understand the durability and evolution of these adhesive-pigment systems. Furthermore, while FTIR provided valuable identification of the collagenous binders, it has inherent limitations in resolving complex protein mixtures or specific degradation pathways. More advanced proteomic methods could offer superior specificity in differentiating glue types and assessing their state of preservation.

This study provides experimental evidence supporting the hypothesis that while all animal glues share a collagenous origin, their specific physicochemical characteristics and concentrations distinctly affect the aesthetic and physical qualities of polychrome coatings. This work reinforces the importance of adhesive selection in cultural heritage conservation, offering insights into historically informed restoration practices.

5. Conclusions

This study comprehensively evaluated the influence of three traditional animal-based adhesives—bone glue, hide glue, and fish glue—on the chromatic, gloss, and microstructural properties of wood-supported polychrome coatings. FTIR confirmed their collagen-based nature, with bone glue showing a stronger hydrogen-bonded band at 3280 cm⁻¹, reflecting compositional differences. Viscosity analysis revealed that bone glue exhibited the highest rheological stability at increasing concentrations, followed by fish glue and hide glue.

Colorimetric results indicated that the glue type and concentration jointly affected the coating appearance. Red cinnabar coatings were primarily influenced by concentration, while yellow orpiment and blue lazurite coatings displayed stronger glue-type effects, with hide glue enhancing chroma and fish glue increasing lightness. Gloss and surface roughness measurements showed that coatings generally retained a matte finish, but bone glue produced smoother and glossier red surfaces, whereas hide glue tended to increase roughness, especially in yellow coatings. Higher adhesive concentrations consistently raised surface irregularities, affecting gloss variability.

These findings demonstrate that while all animal glues share a collagenous origin, their specific physicochemical characteristics and concentrations distinctly shape the aesthetic and structural properties of polychrome coatings. This study provides a scientific basis for selecting historically appropriate adhesives in the conservation of East Asian polychrome artifacts, supporting both visual authenticity and material stability.