Introduction

Textiles, as an important type of cultural artifacts, are valuable physical materials with extremely high research value. The textile artifacts have the characteristics of low density and porosity, and the main components are natural organic fiber, which make them more susceptible to microbial damage and degradation. Therefore, compared to other types of artifacts, the number of preserved textiles is relatively scarce1. Even in storage environments where cleanliness, light, temperature, and air quality are strictly controlled, textile artifacts are still vulnerable to mold attacks. Many museums collect large numbers of textile artifacts, and during long-term preservation, biological corrosion poses a greater risk than chemical corrosion. Therefore, finding suitable decontamination techniques is crucial for the preservation of textile artifacts.

Currently, chemical methods dominate the decontamination of textile artifacts2,3. Common reagents, such as ethylene oxide, are frequently used in these processes4,5,6. Although Chemical reagent fumigation is a very convenient method, it presents significant environmental and health risks. Ethylene oxide not only harms the environment but also fosters microbial resistance7,8,9. Nowadays, radiation decontamination, particularly using gamma rays, has also gained attention10,11. Although it is highly environmentally friendly, the major challenge faced by this method is the expensive facility costs, which include both the initial setup and ongoing maintenance expenses7. Additionally, many artifacts cannot be easily moved to the sterilization facility due to preservation requirements, and this often involves cumbersome paperwork and approval procedures7. Due to the limitations of existing methods for decontaminating textile artifacts12,13, the Electron irradiation with mobile and controllable radiation devices has emerged as a potentially promising method for on-site museum decontamination and deinsectization. This method typically requires lower doses for insect disinfestation compared to sterilization, as the DNA complexity of pests makes them more susceptible to radiation at lower levels, effectively deactivating insect eggs and rendering adults infertile13. In addition, electron beam boasts higher dose rates than gamma rays, enabling swift decontamination of artifacts13. Moreover, electron beams focus energy specifically on target surfaces, enhancing efficacy in comparison to gamma rays. With its unique advantage in surface decontamination, electron beam decontamination proves to be an ideal method for textile artifacts, given their low density and thinness7,14.

While electron beam sterilization is well-established in food and medical applications, its use for textile artifact decontamination is still underdeveloped. There is limited research in this area, and even fewer studies focus on its practical application to textile artifacts. As a result, several challenges remain unresolved7, such as the effectiveness against common microorganisms and the tangible impact on the matrix of real artifacts. Textile artifacts, primarily composed of materials like cellulose and protein, are highly susceptible to damage caused by irradiation-induced free radicals15. For electron beam decontamination, the radiation dose must be calibrated to sufficiently reduce microbial load to a safe level while simultaneously preventing damage to the textile artifacts16. Therefore, the dosage of the electron beam is poised to be one of the most critical parameters in the experiment. When implementing an electron irradiation decontamination plan for specific cultural artifacts, it is crucial to possess a detailed and comprehensive understanding of the decontamination effectiveness and its impact on the textiles. However, there is currently limited research on the effects of irradiation on natural fibers, and available reference materials are scarce. This knowledge is essential for determining the appropriate dosage to be utilized.

In this study, electron beam decontamination experiments were conducted on the four most common fungi found in fibers to validate the feasibility of decontamination. Additionally, performance analysis was carried out on three fibers (cotton, flax, and silk) before and after irradiation. The assessment included parameters such as molecular degree of polymerization, alterations in chemical structure, elongation at break, breaking load, and color. These experiments aimed to comprehensively understand the impact of electron beams on both textile artifacts and microorganisms, thus determining suitable dosages for decontamination across different types of textile artifacts. Our study provides a reference basis for establishing optimal electron beam irradiation procedures for conservationists when sterilizing various textile artifacts.

Methods

Materials

Ancient textile artifacts can be categorized by raw material types, including cotton, flax, silk, and wool. They can also be classified by material composition into animal fibers and plant fibers. All fiber samples used in this study were provided by the Palace Museum. These fibers are natural fibers that have not undergone dyeing or chemical processing treatments, with a history of 30 years. The samples include mulberry silk fibers, flax fibers, and cotton fibers. Given that we are only at the initial exploration stage, we have not yet conducted experiments on actual cultural artifacts.

To assess the effect of radiation dose on microorganism, we exposed four common fungi found in textile artifacts—Aspergillus flavus, Aspergillus niger, Aspergillus versicolor, and Penicillium italicum—to irradiation using single-energy electron beams (200 keV) at various doses17,18,19,20. The four types of fungi (Fig. 1) were revived, cultured for 3–4 days, spot inoculated. Each type of fungus had three culture dishes, with three inoculation points per dish, totaling nine inoculation points for each fungus.

Fig. 1: Four types of fungal colonies without electron beam irradiation.
figure 1

This figure shows four different fungal colonies observed under the same conditions. a Aspergillus niger. b Aspergillus flavus. c Penicillium italicum. d Aspergillus versicolor. All images were taken under the same lighting and magnification conditions.

Full size image

Methods

Considering that decontamination requires radiation with higher surface energy deposition, we use a 200 keV electron beam, which is more suitable for commercialization due to its lower cost. All irradiation was performed using the MEB-200 electron beam irradiation equipment in a nitrogen-protected environment, with the electron beam energy set at 200 keV. The dose rate of the electron beam irradiation equipment was 1.17 kGy/s. Electron beam irradiation was performed on A. niger, A. flavus, A. versicolor, P. italicum, and different kind of fibers using doses of 5, 7.5, 10, 12.5, 15, 20, 25, and 30 kGy, respectively. Irradiation was conducted after inoculation, followed by a five-day incubation period in a culture chamber at 28 °C. The primary premise of electron beam decontamination is that it should not cause significant impacts on the morphological microstructure, mechanical strength, and appearance of textile artifacts. Considering this, we conducted research on the molecular weight and chemical structure of fibers, as well as the elongation at break, breaking load, and color before and after irradiation. This examination aimed to assess the impact of irradiation on the degree of fiber molecular polymerization and molecular structure, as well as the changes in mechanical strength and the effect on fiber color.

Using the Instron Tensile Tester 5942, fracture elongation and fracture load tests were conducted on fibers irradiated with electron beams at different doses (5–30 kGy). The gauge length for a single fiber was set at 20 cm, with a cross-head speed of 200 mm/min. Each type of fiber was divided into nine groups, with 20 samples per group. For each sample, the fracture elongation (%) and fracture load (N) were measured, and the average values were recorded. Measurements were conducted at a humidity of 65% and a temperature of 20 °C. Colorimetric tests were conducted on the fibers before and after irradiation using the 3nh SC-10+ colorimeter. Cotton and flax were dissolved in a copper ethylenediamine solution. The viscosity of the solutions was then measured using a Ubbelohde viscometer, and their degree of polymerization was determined using a characteristic viscometer, following the standards outlined in GB/T 1548-201621. The chemical changes due to electron irradiation in silk were monitored by Fourier transform infrared (FTIR) spectroscopy using a Thermo Scientific Nicolet iZ10 FTIR spectrometer. The spectra were scanned in the range starting from 500 cm−1 to 4000 cm−1. For data processing, we used OMNIC 1.70 software, which was employed for automatic baseline correction and smoothing of the spectra. The data visualization and analysis in this paper were performed using the ROOT framework (version 6.30.06) and Python22.

Results

Decontamination effect of electron beam

Ionizing radiation can interact with microorganisms in two ways: (i) direct interaction with the cell components such as DNA, and (ii) the indirect modification produced by free radicals resulting from water radiolysis. It is the latter indirect effect that is the predominant pathway of inactivation of microorganisms. Important free radicals like hydroxyl radicals (OH·) are formed in the hydration shell of the DNA molecule. They are responsible for 90% of the DNA damage13,23.

For decontamination effectiveness of irradiation, D10 is an important parameter, representing the irradiation dose necessary to reduce the number of microorganisms by a factor of ten13. However, reaching the D10 dose does not guarantee the complete eradication of all bacteria, nor does it signify the reduction of spore germination rates to zero. Consequently, practical decontamination dosages often surpass the D10 threshold. This is evident in our research. For example, A. flavus exhibits a D10 range of 2.0–2.5 kGy at a cultivation temperature of 30 °C under a γ-ray dose rate of 0.12 kGy/h10. However, as depicted in Fig. 2, we observed that more than 5 kGy is required to ensure no spore multiplication. Similar observations were made for other fungi. At 28 °C, A. niger demonstrates a D10 of 0.68 kGy under a γ-ray dose rate of 1.95 kGy/h24 while more than 7.5 kGy is needed in our experiments. Similarly, A. versicolor exhibits a D10 of 0.282 kGy under experimental temperatures between 20 °C and 22 °C and a γ-ray dose rate of 12.2 kGy/h, but only doses exceeding 15 kGy can ensure no colony formation. As for P. italicum, under electron beam irradiation at 1 kGy/s, it shows a D10 of 0.234 kGy, while doses not exceeding 5 kGy are required to prevent colony formation.

Fig. 2: The colony emergence rates of four inoculated fungal colonies after being irradiated by different doses of electron beams.
figure 2

a Aspergillus niger. b Aspergillus flavus. c Penicillium italicum. d Aspergillus versicolor.

Full size image

In this paper, we utilized colony growth rate to assess the effectiveness of electron beam decontamination. The experimental parameters were carefully designed to provide a comprehensive evaluation of the decontamination efficacy. Based on the decontamination performance, it is evident that electron beam irradiation significantly inhibited the growth of four inoculated fungal species. Electron beam irradiation exhibited a substantial decontamination effect on P. italicum and A. versicolor as shown in Fig. 2. Particularly for P. italicum, with a dose of 5 kGy, the spore germination rate was reduced to zero, indicating the electron beam’s effective capability to eliminate the growth of this type of fungus. However, A. versicolor required a higher irradiation dose for complete inhibition. The spore germination rate dropped to zero only at doses exceeding 15 kGy. Nevertheless, as the dose increased, A. flavus and A. niger began to emerge. This could be attributed to the close intertwining of the hyphae of these fungi, leading to excessive biomass at the inoculation points.

One possible reason for the rapid decline in spore germination rate at lower irradiation doses is that when the population of viable cells is high, the probability of interaction with irradiation increases, leading to a more pronounced sterilization effect. However, at higher doses, as the number of surviving cells decreases, the likelihood of irradiation directly affecting the remaining cells is reduced. This may explain why, in our sterilization experiments, certain microorganisms remain difficult to completely eliminate even at elevated doses, with a few colonies still detectable.

Chemical structure changes in fiber after irradiation

Flax and cotton are plant fibers, and cellulose is their main component. The electron beam irradiation of the target material induces the generation of free radicals. These free radicals may lead to the cleavage of secondary structure and covalent bonds, thereby disrupting the molecular structure within the fibers, especially the main chains of large molecules. It is generally accepted that one important effect of irradiation is the cleavage of main chain linkages, which can be detected by a decrease in viscosity25. From the viscosity experiment, we can get the relationship between electron beam irradiation dose and degree of polymerization of cotton and flax as shown in Fig. 3. The molecular weight of cotton and flax fibers exhibits an exponential decrease with increasing dose. However, compared to flax fibers, the molecular weight of cotton fibers decreases more slowly.

Fig. 3: The impact of electron beam irradiation at different doses on the degree of polymerization of cotton and flax fibers.
figure 3

The data points represent experimental results, and the dashed lines represent exponential fitting.

Full size image

In cellulose, the hydroxyl radicals generated by irradiation combine with hydrogen atoms on carbon atoms to form hydroxyl radicals, which subsequently oxidize into corresponding carboxyl groups. The formation of carboxyl groups on carbon atoms C2, C3, and C6 of cellulose may lead to the cleavage of glycosidic bonds, reducing the polymerization degree of cellulose chains13. As shown in Fig. 3, the polymerization degree of both cotton and flax fibers decreases exponentially with increasing dose, but the decline is less pronounced in cotton. Although radiation induces the generation of free radicals in cellulose, resulting in a decrease in the polymerization degree, these free radicals may also undergo cross-linking reactions, leading to an increase in the polymerization degree13,26,27. This may be one reason why the decrease in the breaking load of cotton fibers is not significant. It is worth noting that some studies suggest that certain natural wood components may contain radiation-resistant elements28, which could be another factor contributing to the less pronounced reduction in the molecular weight of cotton fibers.

Due to sample limitations, we only conducted relevant studies on silk, which is an animal fiber. Silk fibers are composed of fibroin enveloped in sericin29. To study the bandwidths of amide I, II, and III, we performed FTIR analysis on silk samples under different irradiation doses. As shown in Fig. 4, four irradiation doses were selected: the control group (0 kGy), 10 kGy, 20 kGy, and 30 kGy. For the β-sheet structure, the peaks of amide I (1632 cm−1), amide II (1550 cm−1), and amide III (1260 cm−1) are marked with dashed lines, while for the random-coil structure, the peaks of amide I (1650 cm−1) and amide II (1515 cm−1) are represented by solid lines30.

Fig. 4: Fourier transform infrared spectroscopy of silk after electron beam irradiation at 0, 10, 20, and 30 kGy.
figure 4

The dashed line represents characteristic peaks of the beta-folded structure (beta), while the solid line represents characteristic peaks of the random-coil structure (RC).

Full size image

When subjected to varying doses of γ radiation, silk fibroin films undergo processes involving the formation of free radicals, the rupture of hydrogen bonds and peptide bonds in the amorphous region, and the disorganization of silk protein chains. The main secondary structure of silk fibroin consists of random-coil and antiparallel β-sheet types, with the β-sheet structure formed through hydrogen bonds (-NH and O = C-) between peptide chains31. When β-sheet structures in fibroin transition to random-coil structures, silk becomes more prone to degradation29,32. It is well established for fibroin and other proteins that the absorption frequency peak position is highly sensitive to the secondary structure31. For random-coil structures, the peaks of amide I and amide II are located at 1650 cm−1 and 1515 cm−1, respectively. In contrast, β-sheet structures have peaks of amide I at 1632 cm−1 and 1695 cm−1, and amide II at 1550 cm−1 30. The peak of amide III is at 1260 cm−1, which is characteristic of β-sheet structures. Regardless of whether it is a random-coil or a β-sheet structure, amide II is present with a peak at 1540 cm−1. If a significant conversion from the β-sheet structure to random coil occurred during irradiation, some peaks corresponding to the β-sheet structure would disappear and be replaced by amide peaks associated with random coil.

However, irradiation at 30 kGy did not cause any β-sheet peaks to disappear, and all peaks appear consistent with those of the non-irradiated sample. Hydrogen bonding is a key component that links the β-sheet secondary structure of silk fibroin. From this observation, we can preliminarily conclude that electron beam irradiation below 30 kGy does not alter the secondary structure of silk fibroin proteins nor significantly disrupt the hydrogen bonds between peptide chains. Similarly, studies on the effects of γ-rays on fibroin suggest that even irradiation doses exceeding 1000 kGy do not alter the primary structure of fibroin33. It is important to note that we only performed automatic baseline correction and smoothing, so the specific changes in peak widths will require further investigation.

Mechanical property changes

The test results for the elongation at break and breaking load of cotton, flax, and silk fibers can be found in Fig. 5. The x-axis in Fig. 5 represents the irradiation dose, the left y-axis corresponds to elongation at break (%) (indicated by the black points), and the right y-axis corresponds to breaking load (N) (represented by the red points). All data points are the average values of 20 parallel tests. The three graphs in the figure correspond to silk, cotton, and flax fibers, respectively. The results indicate that, with an increase in irradiation dosage, the tensile strength of cotton, flax, and silk fibers exhibits a trend resembling exponential decay, aligning with the exponential fitting function proposed by Winogradoff34. However, some studies on silk fibers suggest a linear decrease in breaking elongation and breaking load after γ irradiation35. Additionally, other literature points out that the changes in the mechanical properties of jute fibers before and after irradiation are nonlinear36. Considering the results of this experiment, the fitting curves (dashed lines) in Fig. 5 are all based on exponential fitting.

Fig. 5: The changes in the breaking elongation and breaking load of silk, cotton, and flax fibers after irradiation with different doses of electron beams.
figure 5

The left y-axis represents the breaking elongation (percentage), and the right y-axis represents the breaking load (Newton). The dashed lines in the graph represent the exponential fitting curves. a represents silk fibers, b cotton fibers, and c flax fibers.

Full size image

Below 10 kGy, compared to flax fibers, cotton fibers exhibit better stability in terms of their degree of polymerization under irradiation. After irradiation with different doses of electron beams, the breaking elongation of cotton has a slight decrease, but the breaking load of cotton remains relatively unchanged, possibly due to cross-linking reactions that enhance mechanical integrity. On the other hand, both the breaking load and breaking elongation of flax show a slight decrease. Although there is a downward trend in the mechanical property, considering the experimental errors, the change is quite negligible. This indicates that cellulose has a certain radiation resistance. Below 30 kGy, as the radiation dose increases, the degree of polymerization of plant fiber molecular chains tends to decrease, but the macroscopic mechanical performance changes are not significant. Some studies suggest that the reduced polymerization degree of cellulose chains results in a minor decline in the breaking elongation and breaking load of the fibers32.

In the case of silk, disorganizations of protein chains ultimately result in a reduction in the average molecular orientation. As a consequence, both the breaking strength and breaking elongation of silk fibroin films decrease32. A similar decline is observed in our study, though it remains minimal at electron beam doses below 30 kGy. This limited effect may be due to the short irradiation time caused by the high dose rate, during which the molecular chains do not undergo a significant amount of breakage. This inference is further supported by the minimal changes observed in the secondary structure of silk in FTIR analysis.

Color changes

Color is a key characteristic of all textile artifacts, and to quantify the color changes that may occur because of radiation exposure, the colorimetric parameters were determined using the CIE L*a*b* procedure. The experimental results are shown in Figs. 6 and 7. The changes in L*, a*, and b* values (Fig. 6) can be converted into x and y coordinates on the CIE chart (Fig. 7). For further details on the conversion process, please refer to Bruce Lindbloom.

Fig. 6: The color changes of fibers after irradiation with different doses of electron beams.
figure 6

a–d represent Delta L, Delta a, Delta b, and Delta E of silk fibers, respectively. e–h represent Delta L, Delta a, Delta b, and Delta E of cotton fibers, respectively. i–l represent Delta L, Delta a, Delta b, and Delta E of flax fibers, respectively. The red dots represent the color difference of the fibers after half an hour of electron beam irradiation compared to before irradiation, while the blue dots represent the color difference of the fibers after irradiation and 3 days of storage in a darkroom compared to before irradiation.

Full size image
Fig. 7: The CIE diagrams depicting color changes of fibers after irradiation with different doses.
figure 7

The asterisks represent pre-irradiation colors, circles represent colors within 30 min after irradiation (0 days after irradiation), and squares represent colors after placing in a darkroom for 3 days. a Flax fibers; b cotton fibers; c silk fibers.

Full size image

Figure 6 illustrates the changes in the colorimetric parameters L*, a*, b*, and E for fibers non-irradiated and irradiated with doses of 5, 10, 15, 20, 25, and 30 kGy. The first to third rows represent the color parameters for silk, cotton, and flax fibers, respectively. From Fig. 6a, e, i, we can observe that all three fibers, cotton, flax, and silk, darken to varying degrees after electron beam irradiation (L* decline). From Fig. 6b, f, j, we can observe that all color parameters generally increase with increasing irradiation dose. It should be noted that in Fig. 6, we adopt the following formulas:

$$\begin{array}{lll}\Delta {L}^{* }\,=\,{L}_{{\rm{irradiation}}}^{* }-{L}_{{\rm{before}}}^{* }\\ \Delta {a}^{* }\,=\,{a}_{{\rm{irradiation}}}^{* }-{a}_{{\rm{before}}}^{* }\\ \Delta {b}^{* }\,=\,{b}_{{\rm{irradiation}}}^{* }-{b}_{{\rm{before}}}^{* }\\ \Delta {E}^{* }\,=\,\sqrt{{\left(\Delta {L}^{* }\right)}^{2}+{\left(\Delta {a}^{* }\right)}^{2}+{\left(\Delta {b}^{* }\right)}^{2}}\end{array}$$
(1)

Figure 7 presents the CIE charts for the three fibers after irradiation. Figure 7a–c show the CIE charts for flax, cotton, and silk, respectively, before and after irradiation with doses ranging from 5 to 30 kGy. In the CIE charts, the asterisks represent the pre-irradiation color, the circles represent the color immediately after irradiation at various doses, and the squares indicate the color measured 3 days after storage in a darkroom. From Fig. 7, it is evident that both silk and cotton fibers exhibit a noticeable color shift on the CIE chart following electron beam irradiation. However, the color partially recovers after 3 days of storage in the darkroom. In contrast, there is no significant overall change in the color of flax. While some darkening occurs, the effect of electron beam irradiation is not as pronounced for flax, particularly at doses below 30 kGy.

Research has indicated that γ-ray irradiation causes darkening of colors in cotton, flax, wool, and silk fibers. This darkening effect intensifies with increasing dosage, particularly in the case of dyed fibers16,37. We also observed this darkening phenomenon in our undyed fibers. For cotton fibers, some yellowing occurs after irradiation (with a decrease in b*), especially when the irradiation dose reaches 10 kGy or higher, as shown in Fig. 6g. However, from Fig. 7b, it can be seen that cotton fibers completely recover to the pre-irradiation color when the electron beam dose is below 30 kGy. However, when we focus on flax fibers, the color change is not very significant. The color tones of the flax before and after irradiation in Fig. 7a almost overlap, indicating that electron beam irradiation with doses below 30 kGy does not impact the color of flax fibers. Due to the higher lignin content in hemp fibers, some studies suggest that aromatic compounds with conjugated double bonds can absorb the energy from radiation, thereby reducing its secondary effects38,39,40. Therefore, this could be one of the reasons why flax fibers show minimal color change after irradiation. The discoloration of plant fiber fabrics primarily results from the formation of conjugated double bonds. Over time, these double bonds may react with atmospheric oxygen, leading to the fading or reversal of color changes41. We hypothesize that a similar mechanism explains the observed recovery of color when irradiated samples are stored in the dark, as the reactive free radicals are gradually neutralized.

During the early stages of silk aging, color change is the most sensitive and easily observable indicator among all characteristic indices, making it the earliest sign of silk degradation42. Some literature indicates that under ultraviolet irradiation, certain key amino acids (such as glycine, alanine, tyrosine, and serine) within the large molecules of silk fibroin are oxidized to form carbonyl-containing emission products, which are the primary reason for the yellowing of silk43. Although the secondary structure of the silk protein remains unchanged in FTIR, alterations in side-chain amino acids can lead to changes in protein color. In our study, silk exhibits noticeable yellowing (indicated by an increase in b*) after irradiation, as shown in Fig. 6c. However, similar to cotton fiber, the color of irradiated silk partially recovers after being placed in a darkroom for 3 days. Additionally, some literature suggests that irradiation causes changes in the color of spruce, and that the color can be restored through heating44.

In summary, although irradiation causes varying degrees of darkening in plant fibers, the effect on flax fibers is almost negligible. When the irradiation dose of cotton fibers is below 10 kGy, there is no significant color difference (ΔE* ≤2), even if the dose reaches 30 kGy, it can be restored after being placed in a darkroom. For silk, irradiation with 30 kGy of electron beams does not significantly affect its protein structure or fiber mechanical properties, but it does result in considerable color deviation. This indicates that silk protein also possesses a certain degree of radiation resistance. However, this color deviation diminishes after being placed in a darkroom, especially for irradiation doses below 15 kGy, where the color deviation has no significant change (ΔE ≤ 2).

Discussion

Taken together, our findings suggest that low-energy electron beam irradiation can be applied effectively for the decontamination of textile heritage materials. Based on our results, we summarize the following key points:

  1. 1.

    At doses below 15 kGy, increasing the irradiation dosage, the decontamination effect on A. flavus, A. versicolor, A. niger, and P. italicum becomes more pronounced. However, when the dosage exceeds 15 kGy, there is no significant additional inhibition of colony growth, and complete eradication of certain fungi is not attainable.

  2. 2.

    In the practical application of electron beam irradiation for decontamination, consideration needs to be given not only to the decontamination effect but also to the potential damage to the artifacts. The appropriate decontamination dosage must be tailored according to the type of fiber.

  3. 3.

    Below 30 kGy, cotton, flax, and silk fibers all exhibit a certain degree of radiation resistance. Beyond 10 kGy, significant changes in color occur in cotton fibers, and there is also a considerable decrease in the degree of polymerization in flax fibers. Additionally, above 15 kGy, noticeable changes in the color of silk fibers occur. It is worth noting that the color deviations in cotton and silk fibers resulting from lower dosages can be restored by placing them in a darkroom. However, the color recovery of irradiated textile heritage objects require further comprehensive investigation.

  4. 4.

    Doses below 10 kGy significantly enhance decontamination efficacy without inducing major changes in fiber color, structure, or mechanical properties. Therefore, we recommend a decontamination dose below 10 kGy as optimal for preserving the integrity of cotton, flax, and silk textile artifacts.