Influence of Different Retting on Hemp Stem and Fiber Characteristics Under the East of France Climate Conditions

ABSTRACT

Historically, for fiber extraction, the stems were immersed in water. However, this method proved far too polluting, leading to its substitution by the dew-retting method in the field. Different retting processes and durations can produce significant differences in hemp stems and fiber characteristics. Both dew- and water-retting methods were carried out on the Santhica 27 cultivar in Eastern France in 2020. An analysis of the stem chemical composition and characteristics has been performed to observe the stem degradation during retting for both methods. Secondly, the analyses of the fiber chemical composition, color, morphological and mechanical properties have been performed to evaluate the impact of retting on the fiber quality. The effect of the retting on the stem and the fiber is observed after 1 week for water-retting and after 4 weeks for dew-retting. For dew-retting, the percentages of fine fiber and cellulose in the fiber bundle increase after 1 week, while the stem chemical composition decreases. Unlike in the dew-retting, the percentage of fine fiber and cellulose in the fiber bundle increases after 4 weeks, while the potassium percentage in the stem, fiber brightness, and fiber tenacity and strain decrease. Finally, after 8 weeks, the fiber properties were found to be similar for both methods.

摘要

从历史上看，为了提取纤维，茎被浸泡在水中. 但事实证明，这种方法污染太大，在现场被露水脱胶法取代了. 不同的脱胶工艺和脱胶时间会导致大麻茎和纤维特性的显著差异.2020年，对法国东部的Santhica 27品种进行了露水和水浸处理. 已经对茎的化学成分和特性进行了分析，以观察两种方法在脱胶过程中的茎降解情况. 其次，对纤维的化学成分、颜色、形态和力学性能进行了分析，以评估脱胶对纤维质量的影响.在水脱胶一周后和露水脱胶四周后观察脱胶对茎和纤维的影响. 对于脱胶，纤维束中细纤维和纤维素的百分比在一周后增加，而茎的化学成分减少. 与露水脱胶不同的是，纤维束中细纤维和纤维素的百分比在四周后增加，而茎中钾的百分比、纤维亮度以及纤维韧性和应变降低. 最后，八周后，证明两种方法的纤维性能相似.

KEYWORDS:

• Hemp fiber
• dew-retting
• water-retting
• stem characterization
• mechanical properties

关键词:

• 大麻纤维
• 露水脱胶
• 水脱胶
• 茎特征
• 机械性能

Introduction

The retting stage is crucial for extracting the best fiber quality and yield, considering the ecological and economic aspects. Retting is an external action acting on the stems, facilitating the loosening and separation of the hemp fiber bundles from the shives (Müssig 2010). Over the years, the retting method has evolved to adapt to ecological and economic considerations. During the nineteenth century, the stems were submerged in water (still or running water) for several weeks, then laid on the field to dry (Vivien de Saint-Martin et al. 1834). Now, the principal retting-methods are water- (Abot 2010; Morrison et al. 2000), dew- (Bleuze et al. 2020; Liu et al. 2015a; Mazian et al. 2020), enzymatic- (Evans et al. 2002; Li and Pickering 2008; Sumere and Sharma 1991), chemical- (Song and Obendorf 2006; Tahir et al. 2011) and stem water vapor explosion treatments (Garcia-Jaldon et al. 1998).

Regarding dew-retting, the stems are left on the field in contact with the soil. The dew-retting method utilizes the natural cycles of moisture (dew and rain) and of dryness (sun and wind). This method encourages the proliferation of microorganisms on the stem, facilitating its degradation (Bleuze et al. 2020; Djemiel et al. 2020). Conversely, water-retting involves submerging stem bundles in still or running water. These stems are colonized and fermented by anaerobic bacteria (Abot 2010; Djemiel et al. 2020; Müssig 2010).

Several factors, including duration, weather conditions, cultivar, etc., impact the retting quality. The duration is an essential factor for the retting quality (Liu et al. 2015a). Under-retting represents an insufficient degradation of the matrix components that connect the shives and the fibers (Müssig 2010). Over-retting is the result of an excessive degradation, which adversely affects the fiber quality (Müssig 2010). During dew-retting, an unfavorable climate can harm the fibers, thus complicating the control over their quality (Henriksson et al. 1997; Mazian et al. 2018). The weather changed according to the geographical location, and consequently the timing and amount of the rain changed too (Réquilé et al. 2021). Moisture (rain or dew) directly impacts the outcome of dew-retting. The occurrence of rain at the start (Liu et al. 2015a; Réquilé et al. 2018) or at the end (Mazian et al. 2018) of the retting process can influence the quality of the fibers.

Before decortication, the quality of the hemp stems characterizes the efficiency of retting, as indicated by studies of the stem color (Mazian et al. 2018; Ribeiro et al. 2015), its mass (Bleuze et al. 2020), its diameter and its chemical compositions (Arufe et al. 2021; Liu et al. 2015a). The stem color evolves during dew-retting. Gray marks appear on the stem during the first few weeks, indicating a change in color. These marks represent the proliferation of microorganisms during dew-retting, facilitated by the action of moisture and heat (Mazian et al. 2018; Ribeiro et al. 2015). Subsequently, the gray marks spread from the stem bark to the shiv, leading to changes in the color of the hemp fiber (Mazian et al. 2020) and the hemp shiv darkens (Arufe et al. 2021). In the textile industry, there is a preference for light-colored fibers, as they facilitate easier dyeing (Bouloc 2006).

Few parameters characterize the impact of retting on hemp fiber quality, such as color (Mazian et al. 2018; Ribeiro et al. 2015), fineness (fiber diameter or linear density) (Mazian et al. 2018; Placet, Day, and Beaugrand 2017; Réquilé et al. 2021), chemical compositions (rates of cellulose, hemicellulose and lignin) (Abot 2010; Mazian et al. 2018, 2020; Zhang et al. 2008) and tensile properties (breaking force and elongation, tenacity, Young’s modulus, etc.) (Liu et al. 2015b; Mazian et al. 2018; Réquilé et al. 2021). The fiber diameter decreases during dew-retting (Réquilé et al. 2021), leading to a reduction in the percentage of coarse fibers (Placet, Day, and Beaugrand 2017). The chemical composition of the bast fiber evolves during retting, with the rates of pectin and hemicellulose decreasing (Mazian et al. 2018, 2020). In the initial weeks, for both dew- and water-rettings, the cellulose rate increases. The fiber quality improves when the rates of lignin and hemicellulose are minimal and the rate of cellulose is maximal (Bouloc 2006). The hemp fiber mechanical properties depend on the retting conditions. The breaking strain and the stress increased during dew-retting for the study of Santhica 27 cultivar in 2016 (Mazian et al. 2018), whereas for the study of USO 31 cultivar in 2013 in France, the stress decreased after the first retting-week (Liu et al. 2015b).

Water-retting has been acknowledged as the most efficient and optimal retting method. Despite its efficiency, the environmental pollution it causes has led to it being replaced by dew-retting, which is recognized as an environmentally friendly alternative (Bócsa and Karus 1998; Horne 2020; Tahir et al. 2011). In this study, dew- and water-retting are compared according to stem and fiber characteristics. Additionally, the study focuses on the degradation of the plant matrix by microbial action (dew-retting) and by the hydraulic and bacterial actions (water-retting) duration. Initially, stem morphology (mass and diameter) and stem chemical components (Phosphorus, Potassium, Calcium, Zinc, Manganese, etc.) have been analyzed to observe the degradation of the stem during retting for both methods. The study examines how water maceration in water-retting and the combined effects of rain and sun in dew-retting are influencing the speed of stem degradation. Furthermore, stem degradation may lead to soil or water pollution, contingent upon according to the retting method employed. Therefore, by examining the chemical composition of the stems, it is possible to forecast their potential environmental degradation consequences. Secondly, stems are manually defibered and combed to also analyze the fiber properties. Finally, the colorimetric, morphological and mechanical properties are analyzed during retting to observe the changes in the fiber. Therefore, this article studies the impact of the retting method coupled with duration on the evolution of the fiber properties.

Materials and methods

Raw material cultivation

Hemp (Cannabis sativa L, cv Santhica 27) cultivation and retting were carried out in the East of France (Ammertzwiller/47°41’05.7“N 7°10’25.9”E) by an organic farmer. This cultivation was sown on May 9, 2020, at 50 kg.ha⁻¹ in a clay-silty slaking soil with pH 6.9 (Decker et al. 2022). Beer yeast (pro-biotic) was used to activate the soil life in addition to compost tea. The harvest was on October 7, 2020, with two steps: first the hemp seed was harvested by a threshing machine, then a mower blade machine was used to cut the hemp stems. The meteorological data (temperature and rainfall) were collected from a metrological station based approximately 20 km from the retting site and reported on Figure 1. It was extracted from the website: www.meteociel.fr.

Figure 1. Daily weather conditions (minimum and maximum temperatures and rainfall) during the eight weeks of dew-retting from the cutting in October (W0) to the last retting week in December 2020 (W8).

Dew-retting

The first retting method was dew-retting. The stems were left on the ground after cutting for 8 weeks. The alternation of humidity and dryness helped microorganisms proliferate from the stem bark to the shiv. The microorganisms acted on the degradation of the components of matrices linking the shiv and the fiber (Müssig 2010).

Water-retting

The second retting method was water-retting. After cutting, the stems were randomly selected. Fifty-centimeter-long stems were immersed in cold water (10–11°C) in an opaque container (20 L) with a cover and a bubbler system for 8 weeks. Before immersion, the stem diameter and mass were determined. Water was changed three times per week. After draining, the experimental system and the stems were washed with water. Every week, 20 stems per retting method were selected randomly.

Stem distribution

Figure 2 represents the stem distribution in the different transformation stages until their analysis. After cutting, 25 stems were randomly selected for the without retting reference.

Figure 2. Stem distribution in the different transformation stage.

Two hundred stems were also selected for water-retting, representing the selection each week of 25 stems for 8 weeks. Then, those 200 stems were put in a container with water (2.3 Water-retting). At the same time, 200 more stems were left on the ground for dew-retting (2.2 Dew-retting). Characterization of 25 stems was made after-cutting for water-retting and after-retting and after-drying for both retting methods (section Stem Characterization). Drying was in the greenhouse with a room temperature between 20°C and 30°C, i.e. the optimum temperature for plant growth and hemp drying, for 1 month to correspond with the minimal time for drying in the field. After the last stem characterization, the stems were separated into three groups shown in Figure 2. Five stems for the stem chemical composition, 5 stems for fiber bundle chemical composition and 10 stems for the fiber morphological and mechanical properties. A part of the stems was dried in the oven at 75°C, then crushed and chemically analyzed (section Stem Chemical Composition Analysis). Stems for the fiber bundle composition were peeled, dried in the oven at 75°C according to the standard NF EN 13040, crushed and analyzed according to the standard XP U44-162 (2.7 Fiber bundle chemical composition analysis). Finally, the stem for the fiber characterization was peeled to separate the shives and the hemp bundles fiber. Then, the bundle fiber was combed to separate the technical fiber between them. The technical fiber color was analyzed by NF ISO CIE 11664 standard (section Technical Fiber Colorimetric Parameters Analysis). Then, the technical fibers were combed until the extraction of refined fiber. The linear density of the refined fiber was between that of ultimate and technical fibers. Then, linear density measurement and the tensile test were analyzed using the NF EN ISO 1973 and NF EN ISO 2062 adaptation standards (section Linear Density and Tensile Test Analysis).

Stem characterization

In order to evaluate the looseness of stem matter during retting, the stem characterization was being determined. The mass was determined with the Mettler Toledo PC400 weighing scales (±0.01 g), and the diameter of the middle of stem was measured with a digital caliper (±0.01 mm). The characterization was made on twenty-five 50-cm-long hemp stems per method and per duration. The mass and diameter were determined after-cutting for water-retting, and after-retting and after-drying for both methods. Two hundred samples were represented from different retting methods and measurement stages based on data from multiple retting durations. Whatever the method, particularly after water-retting, the stems were drained in ambient air before mass and diameter determination. Draining was long enough to consider the stem to be free of water. The stem masses were classified from higher to smallest smaller. Then, the evaluation of the water moisture absorption rate was calculated by the following equation:

$\mathit{H}\mathrm{\%}=\frac{{\mathit{M}}_{\mathit{retting}}-M_{\mathit{dry}}}{M_{\mathit{retting}}}\times 100$

H% representing the water moisture rate, M_retting representing the hemp stem mass after-retting and M_dry representing the mass of the hemp stem after drying in a greenhouse whose temperature represents the temperature of the field.

Stem chemical composition analysis

The chemical composition of the hemp stem during the retting degradation was analyzed by SADEF laboratory. Only one sample, consisting of five stems, per retting method and duration was analyzed. The hemp stems were dried at 75°C in an oven to eliminate the water without degrading the stem chemical composition for a time long enough to consider the stem free of water (around 1 week). Then, the stems were crushed. For the analysis of chemical composition, moisture loss at 105°C during 12 h was achieved on 1 g of vegetal powder. The calcination of the powder consisted of placing the sample in a cold oven, while the temperature was gradually increased for 4 h up to 550°C and then the temperature was maintained at 550°C for 3 h. Finally, the ashes were cooled. The solubilization of the ashes was carried out. The ash was slightly wet (2 ml) with mineralized water, and a few drops of hydrochloric acid 32% were added. Then, 5 ml of hydrochloric acid 32% was added with an OPTIFIX dispenser. The sample was heated on a plate at 150°C until evaporation. Finally, 10 ml hydrochloric acid 60% was added to the ash. Filtration was performed using a funnel, an ash-free filter and a filtration ramp. The filters were then rinsed with 10% hydrochloric acid and then twice with demineralized water. Finally, the quantitative analyses were carried out with ICP MS (Inductively Coupled Plasma-Mass Spectrometry) named THERMO Serie3®. Then, the chemical composition rate was calculated with the following equation:

$\mathit{X}\left(\mathit{mg}\right)=0.01\frac{C}{{\mathit{M}}_{\mathit{initial}}}\times \frac{100}{{\mathit{M}}_{\mathit{dry}}}\times m_{\mathit{stem}}$

$\mathit{X}\left(\mathit{\mu g}\right)=100\frac{C}{{\mathit{M}}_{\mathit{initial}}}\times \frac{100}{{\mathit{M}}_{\mathit{dry}}}\times m_{\mathit{stem}}$

X representing the element rate, C representing the element rate in mg/L, M_initial representing the mass of the sample before test, M_dry representing the mass of the sample after the moisture loss at 105°C during 12 h and m_stem representing the mass of the stem previously determined.

Fiber bundle chemical composition analysis

Cellulose, hemicellulose, soluble fraction and lignin analyses were carried out on cut and crushed bundle fiber according to the Van Soest method (Abot 2010) and the FD U44-162 standard. Only one sample per retting method and duration was analyzed. The small piece of bundle fiber had a succession of three acid attacks. The first attack by Neutral Detergent to extract neutral detergent fiber (NDF) consisted of the soluble fraction (pectin and protein). The second attack by Acid Detergent to extract the acid detergent fiber (ADF) consisted of hemicellulose and cellulose. The third attack using sulfuric acid to extract the acid detergent lignin (ADL) consisted of the lignin. Finally, the sample had a calcination to calculate the mineral rate. Then, the cellulose and hemicellulose rates were calculated by the following mass difference (Angelini et al. 2014; Jankauskienė et al. 2015):

$\mathit{Cellulose}=\mathit{ADF}-\mathit{ADL}$

$\mathit{Hemicellulose}=\mathit{NDF}-\mathit{ADF}$

ADF representing the proportion of acid detergent fiber, ADL representing the proportion of acid detergent lignin and NDF representing the proportion of neutral detergent fiber.

Technical fiber colorimetric parameters analysis

The fiber color changed during dew retting (Akin et al. 2000; Bleuze et al. 2018; Epps et al. 2001; Martin et al. 2013; Müssig 2010). According to the NF ISO CIE 11664 standard, a spectrophotometer named Datacolor 500® was used. This instrument measured in the color space system L*a*b CIE 1976. The color was defined by three chromatic parameters L* (whiteness/darkness), a* (redness/greenness) and b* (yellowness/blueness). The instrument calibration was made with a light trap and standard white and green colors. Five measurements were made per sample. The measurement consisted of placing a technical fiber between the hole and the clamp. A white ray is sent on the sample. The instrument analyzed the reflected ray and translated it into chromatic parameters L*, a* and b * .

Linear density and tensile test analysis

The fiber bonds were degraded by the retting action. Consequently, the fiber morphology and mechanical properties were modified (Liu et al. 2015a; Mazian 2018; Placet, Day, and Beaugrand 2017). Refined fiber was selected randomly and placed on two adhesive stickers. According to the NF EN ISO 1973 standard, the linear density was carried out by vibroscopic method. It corresponded to the measure of the fiber resonance frequency with a constant pretension and a fixed distance between the two fiber supports. The sample was placed on the VIBROMAT Me® instrument to measure the approximate linear density with a 0.5 cN.tex⁻¹ pretention. Then, the sample was put in the FAVIMAT +’s clamps. FAVIMAT + instrument® measured the linear density and also the mechanical properties (Decker, Drean, and Harzallah 2021). Its setup was a 20 mm clamp distance, a 210 cN load cell, a 1 mm.min⁻¹ constant crosshead displacement rate and a 0.5 cN.tex⁻¹ pretention configured by an approximate linear density. Firstly, the linear density was measured and classified into seven groups. Secondly, according to the NF EN ISO 2062 standard adaptation, the mechanical properties (tenacity, breaking strain and Young’s modulus) of the sample were performed.

Statistical analysis

The data of the stem characterization, technical fiber color, linear density and mechanical properties were analyzed using one-way analysis of variance (ANOVA). This analysis was conducted using R® software. Other data, such as the chemical composition of the stem and the fiber, were analyzed with machine accuracy.

Results and discussion

Weather condition

The daily weather conditions from the dew-retting are presented in Figure 1. The rainfall was very significant during the initial weeks, especially between week 0 and week 1 (16.4 mm), between week 2 and week 3 (43.0 mm), and between week 3 and week 4 (11.8 mm). Between week 4 and week 8, there was a noticeable decline in both temperature and rainfall, until the end of retting. Then, the minimum temperatures dropped below 0°C, so probably the microorganism activities were slowed. These retting weather conditions corresponded to a normal retting condition up to week 6.

Stem diameter and mass

Retting impacted the characteristics of the stem. Given that the stem length was constrained by the container size, evaluations were limited to the diameter and the mass. Table 1 displays the average values across different stages of all stem diameters and masses data collected throughout an 8-week retting period. Regardless of the retting method, the diameter did not evolve between after-retting and after-drying. However, for water-retting, the stem diameter decreased and the mass increased between after-cutting and after-retting, respectively, from (7.86 ± 0.29) mm to (7.22 ± 0.27) mm and from (9.57 ± 0.76) g to (20.80 ± 1.56) g. The decrease in diameter results from separating of the bast from the shiv, a process facilitated by bacteria like “Clostridium” that break down pectin, which is the connector between the fiber bundles and the shiv (Djemiel et al. 2020). The increase in mass is attributed to the hemp stem’s absorption of water, leveraging its porous nature (Garcia-Jaldon et al. 1998). The difference observed between the two methods after-drying resulted from the shedding of the bast fibers during the water-retting process.

Table 1. Stem characteristics and confidence intervals estimated for different retting methods and stages based on data from multiple retting durations.

Method	Stage	Diameter (mm)	Mass (g)
Dew	After-retting	7.65 ± 0.23 ab	6.68 ± 0.41 c
	After-drying	7.43 ± 0.19 ab	4.66 ± 0.26 d
Water	After-cutting	7.86 ± 0.29 a	9.57 ± 0.76 b
	After-retting	7.22 ± 0.27 bc	20.80 ± 1.56 a
	After-drying	6.82 ± 0.24 c	3.69 ± 0.29 e
p-value		<0.001	<0.001

The letters a to d correspond to the comparison of the impact of the method coupled to the drying in the same column on the stem characteristic according to an ANOVA analysis.

Regardless of the method used, the mass significantly decreased after-drying, in particular for water-retting from (20.80 ± 1.56) g to (3.69 ± 0.29) g. This can be explained by the fact that the amount of water present during dew-retting is lower and more variable than that encountered during water retting, according to Figure 1. So, hemp stems contained water. The moisture rate quantified the water composition of the stems. Figure 3 represents the amount of water absorbed during different weeks of retting, depending on the method used. The amount of water absorbed during dew-retting, around 2 g (30%), was significantly lower than the amount during water-retting, around 17 g (82%), according to Figure 3. It changed during retting, because the weather was not constant, as shown in Figure 1. Between weeks W2 and W3, the moisture rate of the stem increased from (13.6 ± 0.6) % to (44.2 ± 1.4) %, which corresponds to heavy rainfall of 43.0 mm. The rate remained stable between W5 and W7, as no rain fell and temperatures cooled down. Between weeks 6 and 7, the weather was dry and cold with a temperature ranging between −5°C and 10°C and no rain, compared to the beginning of the retting. The moisture rate in week 7, (16.97 ± 1.36) %, was lower than in week 6, (27.72 ± 1.59) %, according to Figure 3. So, the moisture rate in dew-retting depended on the weather conditions, particularly the amount of rainfall. However, the moisture rate remained constant during water-retting. The quantity of water absorbed impacted the development of the microorganisms on the stem during dew-retting. The stem’s moisture rate affected the scutching efficiency. Scutching became increasingly challenging with higher stem moisture, leading to more significant fiber losses. Therefore, it is essential to perform scutching on dry stems to minimize these losses.

Figure 3. Impact of the retting method and duration on the hemp stem moisture rate.

The water moisture rate in the stem fluctuated throughout the retting process. As a result, the following analyses were conducted only on stems that had been dried. Table 2 shows the variations in stem diameter and mass observed with both retting methods. The diameter of the stem using dew-retting stem remained constant, while those using water-retting method fluctuated during retting. This changes in diameter resulted from the loss of fiber bundles during retting, which also explained the fluctuations of the hemp stem masses. The variation in the mass of the stem using dew-retting was lower than that of stem using water-retting. The mass of the water-retted stem decreased between week 1 (5.06 ± 1.14) g to week 8 (3.70 ± 0.90) g, according to Table 2. In contrast to the finding of Ruan study on flax fiber in Canada, which observed a gradual decrease in stem mass over 10 days, the decline in mass during the retting process in this case was not steady (Ruan et al. 2015).

Table 2. Stem characteristics and confidence intervals estimated for different retting methods and durations based on data from after-drying stage.

Method	Week	Diameter (mm)	Mass (g)
Dew	W1	7.62 ± 0.38 a	5.11 ± 0.66 ab
	W2	7.78 ± 0.52 a	5.31 ± 0.78 a
	W3	7.37 ± 0.46 a	4.25 ± 0.57 ab
	W4	7.05 ± 0.60 ab	4.12 ± 0.71 ab
	W5	7.68 ± 0.63 a	5.36 ± 0.99 a
	W6	7.73 ± 0.71 a	5.15 ± 0.96 ab
	W7	6.83 ± 0.57 ac	3.66 ± 0.58 ac
	W8	7.27 ± 0.50 a	4.14 ± 0.72 ac
Water	W1	7.94 ± 0.81 a	5.06 ± 1.14 a
	W2	7.27 ± 0.59 a	4.22 ± 0.73 ab
	W3	5.81 ± 0.55 bc	2.88 ± 0.62 bc
	W4	6.99 ± 0.48 ac	4.08 ± 0.75 ac
	W5	5.59 ± 0.54 c	2.32 ± 0.43 c
	W6	6.81 ± 0.74 ac	3.51 ± 0.79 ac
	W7	6.96 ± 0.73 ac	3.72 ± 0.86 ac
	W8	7.22 ± 0.73 a	3.70 ± 0.90 ac
p-value		<0.001	<0.001

The letters a to c correspond to the comparison of the impact of the method coupled to the duration in the same column on the stem characteristic according to an ANOVA analysis.

Chemical composition analysis of hemp stem

Table 3 presents the masses of the major elements and the oligo-elements in the stem. Prior to retting, the masses of the elements in our study were similar to those of other studies, whereas our Phosphorus mass was slightly different (Hakala et al. 2009; Kleinhenz et al. 2020; Väisänen et al. 2019). This indicates that the development of the hemp stems was optimal before mowing.

Table 3. Stem chemical composition and machine accuracy estimated for different retting methods and durations.

Method	Week	Potassium (mg)	Calcium (mg)	Phosphorus (mg)	Magnesium (mg)	Boron (µg)	Copper (µg)	Manganese (µg)	Zinc (µg)
	W0	206.6 ± 16.6	210 ± 16.9	24.1 ± 1.9	22.4 ± 1.8	232.4 ± 23.4	94.7 ± 14.3	588.8 ± 88.7	172.2 ± 25.9
Dew	W1	263.1 ± 21.2	313.3 ± 25.2	45.4 ± 3.7	23.9 ± 1.9	377.9 ± 37.9	120.5 ± 18.1	1171.9 ± 176.3	73.7 ± 11.1
	W2	169.1 ± 13.6	233.8 ± 18.8	39.8 ± 3.2	17.4 ± 1.4	271.1 ± 27.2	129.8 ± 19.5	614.4 ± 92.4	94.0 ± 14.1
	W3	29.9 ± 2.4	123.4 ± 9.9	13.9 ± 1.1	10.0 ± 0.8	144.7 ± 14.5	73.9 ± 11.1	312.5 ± 47.0	65.3 ± 9.8
	W4	38.6 ± 3.1	250.6 ± 20.2	34.7 ± 2.8	15.4 ± 1.2	219.7 ± 22.1	83.8 ± 12.6	857.8 ± 129.1	74.6 ± 11.2
	W5	45.2 ± 3.6	306.6 ± 24.7	47.7 ± 3.8	15.1 ± 1.2	261.3 ± 26.2	112.3 ± 16.9	1223.7 ± 184.0	105.0 ± 15.8
	W6	36.2 ± 2.9	205.1 ± 16.5	21.7 ± 1.7	12.1 ± 1.0	185.3 ± 18.6	83.2 ± 12.5	460.9 ± 69.3	89.0 ± 13.4
	W7	17.1 ± 1.4	171.5 ± 13.8	13.7 ± 1.1	8.6 ± 0.7	138.5 ± 13.9	97.4 ± 14.7	399.5 ± 60.2	72.9 ± 11.0
	W8	44.6 ± 3.6	286.8 ± 23.1	31 ± 2.5	15.5 ± 1.2	244.2 ± 24.5	108.9 ± 16.4	730.6 ± 110	137.8 ± 20.7
Water	W1	33.2 ± 2.7	284.3 ± 22.9	14.2 ± 1.1	11.8 ± 0.9	241.6 ± 24.3	402.7 ± 60.6	689.3 ± 103.7	1075.4 ± 161.8
	W2
	W3	3.5 ± 0.5	102.1 ± 8.2	3.5 ± 0.3	3.5 ± 0.3	123.8 ± 12.5	440.3 ± 66.3	225.4 ± 33.9	1549.8 ± 233.4
	W4	8.5 ± 0.7	169.9 ± 13.7	18.9 ± 1.5	13.2 ± 1.1	190.6 ± 19.2	556.8 ± 83.8	479.4 ± 72.2	1457.2 ± 219.4
	W5	1.5 ± 0.5	62.5 ± 5.0	1.5 ± 0.1	1.5 ± 0.1	72.3 ± 8.8	252.8 ± 38.1	93.1 ± 17.5	758.3 ± 114.3
	W6	7.8 ± 0.6	79.8 ± 6.4	3.9 ± 0.3	1.9 ± 0.2	103.7 ± 11.7	324.9 ± 48.9	187.4 ± 28.2	953.4 ± 143.5
	W7	1.6 ± 0.5	54.7 ± 4.4	1.6 ± 0.1	1.6 ± 0.1	68.6 ± 9.7	421.8 ± 63.5	75.3 ± 19.4	1001.4 ± 150.8
	W8	3.7 ± 0.6	95.0 ± 7.7	3.7 ± 0.3	1.9 ± 0.2	114.0 ± 11.5	897.5 ± 135.1	191.8 ± 28.9	1845.2 ± 277.8

The process of soaking stems in water (water-retting) accelerates the degradation of Potassium within the stem, due to its high solubility in water. This mass decreased significantly in the first week of water-retting 33.2 ± 2.7 mg and until the third week for dew-retting 29.9 ± 2.4 mg per stem. Subsequently, the mass of Potassium remained constant, independently of the method used. The mass of Calcium within the stem using dew-retting was higher than that within the stem using water-retting. In the process of water-retting, the Calcium mass increased during the first week, and then it decreased from 284.3 ± 22.9 mg week 1 to 95.0 ± 7.7 mg per stem in week 8, because this element is soluble in water.

Phosphorus mass decreased during the first week of water-retting and then remained constant at around 2–4 mg per dry matter. Across both methods of retting, there was a reduction in the mass of Magnesium. In the process of dew-retting, the masses of Phosphorus, Potassium, Calcium and Magnesium were higher as compared to the process of water-retting, attributed to their solubility in water. A similar disparity in the mass of phosphorus, potassium, and magnesium in the stem between the two retting methods was observed in flax stems, as documented by Sharma et al. (1999).

Excluding the data from the fourth week for both methods, Phosphorus and Magnesium showed identical patterns. Their mass in the hemp stems was decreased by the presence of water. In the case of dew-retting, the moisture rate seemed to impact their mass the initial 3 weeks. Following this period, a concentration effect was noted due to the solubilization by water of other elements. Additionally, in dew-retting, significant rainfall occurred from W2 to W4, followed by a decrease until W7. The absence of rain, and consequently the lack of moisture in the stems, prevented their chemical composition from degrading.

During dew retting, the masses of Boron, Copper and Manganese in the stem remained unchanged, as shown in Table 3, implying that these elements did not contribute to soil enrichment. Conversely, the mass of Zinc diminished in the first week of dew-retting, signifying a contamination level of 98 µg per stem. Excessive Zinc levels can lead to soil contamination. Given that the average yield of a hemp crop stands at 8 t/ha and each stem weighs 5 g, dew- retting results in the release of approximately 157 g of zinc per hectare. Considering that a hectare contains 3000t of soil, the increase in zinc concentration post-retting would be 52 mg/t, a negligible increment. Thus, dew-retting has a minimal impact on soil pollution.

This is in contrast to water-retting, where the masses of Zinc and Copper increased over time, indicating potential contamination with Copper and Zinc due to water. The masses of Boron and Manganese decreased during water-retting from 241.6 ± 24.3 µg (W1) to 114.0 ± 11.5 µg (W8) and from 689.3 ± 103.7 µg (W1) to 191.8 ± 28.9 µg (W8), respectively. This reduction is attributed to the solubility of these elements in water. It appears that the water was contaminated with Boron and Manganese, at concentrations of approximately 6 µg and 25 µg per plant per liter, respectively. Nonetheless, these quantities are considered to be negligible.

The masses of Copper and Zinc in the stem using water-retting were higher than those in the stem using dew-retting, attributed to their presence in the tap water. Conversely, the masses of Boron and Manganese were greater in the case of dew-retting attributed to the solubility of these elements in water. Changes in moisture rate did not impact the mass of oligo-elements within the stem. Zinc and copper are mainly absorbed by the stems, and the proportion in the fibers is low. According to the study by Angelova et al. (2004), the hemp fiber would absorb about 60% less than the whole stem. Furthermore, the presence of heavy metals such as Cadmium and Lead does not hinder the quality of the fibers. It can be assumed that Zinc and Copper will not alter the quality of hemp fibers (Citterio et al. 2003; Linger et al. 2002).

Fiber bundle chemical composition

Table 4 presents the chemical composition of the fiber bundle, which is a key in determining the efficiency of decortication, according to Bouloc (2006). In the first week, and particularly in the process of water-retting, the cellulose rate increased from 76.5% (W0) to 85.6% (W1), a finding that aligns with observations made in Abot’s research on water retting (Abot 2010). As noted by Bouloc (2006), the rise in cellulose rate resulted in enhanced fiber quality. After the first week, this rate stayed stable until the end of the retting process (Abot 2010; Mazian et al. 2019). The retting method appears to have a minimal effect on the cellulose rate in fiber bundles, regardless of how long the retting process lasts, as shown in Table 4 and supported by Różańska et al. (2023). On the other hand, findings from Jankauskienė et al. (2015) indicate a higher cellulose rate in fibers subjected to water-retting compared to those from dew-retting. This suggests that factors associated with the retting process, including the temperature of the water and the surrounding environmental conditions, play a significant role in determining the cellulose rate within the fiber bundles. In the end, the cellulose rate was higher than 80% for both methods, which is in line with the study of Abot (2010), whereas the study of Mazian et al. (2019) showed a cellulose rate of 70–75%. The rate of cellulose indicates that the quality of the technical fiber meets the requirements of the textile industry, as stated by Bouloc (2006).

Table 4. Fiber chemical composition and machine accuracy estimated for different retting methods and durations.

Method	Week	Soluble fraction (%)	Hemicellulose (%)	Cellulose (%)	Lignin (%)
	W0	0.0 ± 0.0	4.0 ± 0.6	76.5 ± 11.5	19.5 ± 2.9
Dew	W1	3.8 ± 0.6	0.9 ± 0.1	78.5 ± 11.8	16.7 ± 2.5
	W2	6.3 ± 0.9	6.8 ± 1	78.5 ± 11.8	8.4 ± 1.3
	W3	4.5 ± 0.7	6.1 ± 0.9	81.3 ± 12.2	8.2 ± 1.2
	W4	1.1 ± 0.2	4.5 ± 0.7	83.9 ± 12.6	10.6 ± 1.6
	W5	3.2 ± 0.5	6.1 ± 0.9	83.6 ± 12.5	7 ± 1.1
	W6	3.2 ± 0.5	5.2 ± 0.8	86.2 ± 12.9	5.5 ± 0.8
	W7	5.2 ± 0.8	2.8 ± 0.4	83.6 ± 12.5	8.4 ± 1.3
	W8	4.1 ± 0.6	5.6 ± 0.8	81.1 ± 12.2	9.2 ± 1.4
Water	W1	1.4 ± 0.2	2.6 ± 0.4	85.6 ± 12.8	10.3 ± 1.5
	W2	7.2 ± 1.1	4.5 ± 0.7	81.7 ± 12.3	6.7 ± 1
	W3	5.7 ± 0.9	5 ± 0.8	84.8 ± 12.7	4.5 ± 0.7
	W4	0.3 ± 0.04	2.6 ± 0.4	85.6 ± 12.8	11.5 ± 1.7
	W5	5 ± 0.8	4.6 ± 0.7	86.3 ± 12.9	4.1 ± 0.6
	W6	3.4 ± 0.5	5.1 ± 0.8	88.3 ± 13.2	3.2 ± 0.5
	W7	4.1 ± 0.6	4.2 ± 0.6	86.9 ± 13	4.7 ± 0.7
	W8	2.5 ± 0.4	2.6 ± 0.4	84.2 ± 12.6	10.6 ± 1.6

From W1 to W2, the chemical composition of the fiber showed similar behavior across both methods, with a decrease in lignin rate and an increase in hemicellulose and soluble fractions. Subsequently, from W2 to W4, the hemicellulose rate fell, which was aligned with a reduction in the mass of Calcium in the stem. This is significant as calcium plays a crucial role in the composition of the cell walls’ middle lamella. Consequently, the degradation of calcium in the stem likely impacted the hemicellulose concentration within the fiber bundles. In both retting methods, the sample from W4 was inconsistent with the rest, probably due to an issue during sample handling. From W4 to W6, for both methods, the soluble fraction rate decreased, similar to the reduction in Potassium, Calcium and Phosphorus within the stem, given that Calcium stabilizes pectin and soluble protein, which make up the soluble fraction (Mazian 2018). In the process of dew-retting, there was a decrease in the lignin rate, while in the process of water-retting, there was an increase in the hemicellulose rate. Finally, up to the end of the retting process, there was an increase in both the lignin rate within the fiber and the calcium within the stem (Abot 2010; Liu et al. 2015a), with a decrease in the soluble fraction observed in both methods. In the end, the rate of lignin remained constant regardless of the retting method (Jankauskienė et al. 2015; Różańska, Romanowska, and Rojewski 2023). The soluble fraction and hemicellulose rates of dew-retting were twice as high as those of water-retting.

In both water and dew retting processes, specific microorganisms were crucial for breaking down pectin, lignin and other substances. Furthermore, the observation was that bacterial growth and activity increased in water, amplifying their role in degradation. As a result, the breakdown of the mineral content in both stems and fiber bundles proceeded more rapidly in water than in field conditions. In addition, pectin serves as a bonding agent for fibers and is considered an effective indicator of retting progress. Insufficient removal of pectin indicates that the stem is under-retted. Conversely, excessive degradation of pectin and other materials signals over-retting, resulting in compromised fiber quality (Müssig 2010). When the pectin material decreases, the fiber bundles become finer, leading to an enhancement in defibrability efficiency (Bouloc 2006). Moreover, industrial techniques such as scutching and spinning remove pectin mechanically. So, the chemical composition of the fibers undergoes changes throughout the textile manufacturing process.

Stem and fiber color

The moisture rate and the retting method impacted the color of the stem and the fiber over time. Observations of these color variations in stem bundles and technical fibers were made through, visual inspection and spectrophotometric analysis reciprocally. Table 5 details the impact of the retting method and duration on the fiber color parameters. In the initial week, the stem transitioned from its green color to a light brown color due to the chlorophyll breakdown, in accordance with the study of Mazian (2018). From W2 to W3 of dew-retting, the stem color changed visually, some gray marks appeared, a phenomenon also documented in the literature (Mazian et al. 2018). The color of the fiber was also altered, irrespective of the color parameters outlined in Table 5. This shift in color for both the stem and fibers is associated with an increase in the water moisture rate of the stem, as shown in Figure 3. This phenomenon occurs because microorganism activity is heightened in moist conditions, leading to accelerated microbial growth on the stem surface (Djemiel et al. 2020; Mazian et al. 2018; Ribeiro et al. 2015). Concurrently, the masses of Boron, Calcium, Copper, Magnesium, Manganese and Potassium decreased (Table 3). Between W3 and W4 of dew-retting, there was no significant change in the fiber color and moisture rate, whereas the masses of major and oligo elements increased. After W4, an increase in the microbial community, including fungi and bacteria, caused the stem to darken in the dew-retting process, though the fiber color remained unchanging (Mazian et al. 2018; Ribeiro et al. 2015). At the same time, there was a progressive decrease in the moisture rate and the masses of Boron, Calcium, Magnesium and Manganese. In contrast, the masses of Copper and Potassium in the stems subjected to dew-retting were constant (Table 3). In the final week of dew-retting, there was no significant change in fiber color. However, there was a significant increase in the moisture rate and the masses of Boron, Calcium, Copper, Magnesium, Manganese and Potassium.

Table 5. Technical fiber color parameters and confidence intervals estimated for different retting methods and durations.

Method	Week	L*	a*	b*
	W0	71.9 ± 0.9 a	3.7 ± 0.2 ab	28.2 ± 0.8 a
Dew	W1	71.1 ± 1.0 ab	3.5 ± 0.4 bc	27.9 ± 0.6 a
	W2	70.4 ± 1.6 ac	3.6 ± 0.3 b	27.4 ± 0.5 ab
	W3	54.4 ± 1.4 d	1.8 ± 0.1 f	18.2 ± 0.3 f
	W4	49.2 ± 0.4 f	2.1 ± 0.1 f	17.0 ± 0.5 gh
	W5	54.1 ± 0.6 d	2.1 ± 0.1 f	18.0 ± 0.3 f
	W6	50.1 ± 0.6 ef	1.7 ± 0.1 f	16.3 ± 0.5 h
	W7	51.6 ± 1.5 e	2.1 ± 0.1 f	17.7 ± 0.7 fg
	W8	50.0 ± 1.8 ef	1.8 ± 0.1 f	16.8 ± 0.6 gh
Water	W1	71.1 ± 0.6 ab	3.2 ± 0.3 cd	26.2 ± 0.7 c
	W2	69.6 ± 0.6 bc	2.9 ± 0.2 de	24.2 ± 0.7 de
	W3	69.1 ± 0.8 c	2.7 ± 0.2 e	23.7 ± 0.4 e
	W4	71.7 ± 0.8 a	2.8 ± 0.5 e	24.8 ± 0.6 d
	W5	70.9 ± 0.4 ab	4.1 ± 0.2 a	27.5 ± 0.6 ab
	W6	71.7 ± 0.4 a	3.6 ± 0.1 bc	26.9 ± 0.2 bc
	W7	71.3 ± 0.6 ab	3.3 ± 0.3 cd	26.1 ± 0.3 c
	W8	70.2 ± 0.4 ac	3.6 ± 0.2 bc	26.6 ± 0.5 bc
p-value		<0.001	<0.001	<0.001

The letters a to h correspond to the comparison of the impact of the method coupled to the duration in the same column on the color parameter according to an ANOVA analysis.

In the water-retting process, there was no notable visual change in the color of the hemp stem, while there were changes in the fiber color parameters. Between W1 and W4 of water-retting, the fiber color and the masses of Boron, Calcium, Magnesium, Manganese and Potassium decreased (Table 3). Conversely, the masses of Copper in the stem increased and the moisture rate remained constant. Brightness (L*) increased between W3 and W4, a trend also observed in the flax stem (Ruan et al. 2015). Then, between W4 and W5, the color parameters (a* and b*) increased, while the masses of Boron, Calcium, Copper and Manganese decreased and the masses of Magnesium and Potassium were constant. In the final weeks of the water-retting process, the fiber color, moisture rate and masses of Boron, Calcium, Copper, Magnesium, Manganese and Potassium remained steady.

In the initial 2 weeks of the retting process, the brightness L* was comparable between fibers using dew-retting and water-retting, i.e. around 70/100. Conversely, the dew-retted fibers displayed higher values for the color parameters a* and b*. The observed difference suggested that fibers subjected to water retting lost color without becoming lighter, in contrast to fibers using dew-retting. This loss of color was associated with decreases in both the mineral and chemical compositions of the stems, as well as discoloration of the stems. After W3, all color parameters of water-retting fiber were higher, according to Table 5. Consequently, the fiber using the dew-retting appeared darker and bluer compared to the fiber using the water-retting, a phenomenon similarly noted in flax fiber by Akin et al. (2000). In comparison with flax retting fiber (Akin et al. 2000), hemp fiber processed dew-retting was darker, greener and yellower. Conversely, hemp fibers subjected to water-retting were lighter, redder and bluer than flax fibers subjected to water-retting (Akin et al. 2000).

In summary, factors such as the moisture content of the stem, the chosen retting technique and the duration of the process significantly influenced the color outcomes. There was a correlation between the chemical composition and the fiber color, especially evident at the onset of the retting phase. During the initial weeks, the fibers lost color and the stem chemical composition (Calcium, Potassium, Phosphorus, Manganese and Boron) decreased. Given that fibers retted in water are lighter and closer to white compared to those retted in the field, these fibers are easier to dye in the textile and furnishing industry.

Fiber characterization

The fiber quality is characterized by its fineness, which includes the average of linear density and its distribution, as well as its mechanical properties, such as tenacity, breaking strain and Young’s Modulus, as depicted in Figure 4 and Table 6. There was no change in Young’s Modulus during the retting process, regardless of the method used, a finding that aligns with observations made by Réquilé et al. (2021).

Figure 4. Impact of the retting method and duration on the fiber linear density distribution.

Table 6. Fiber characteristics and confidence intervals estimated for different retting methods and durations.

Method	Week	Linear density (dtex)	Tenacity (cN/tex)	Breaking strain (%)	Young’s Modulus (cN/tex)
	W0	18.08 ± 2.22 b	68 ± 7 a	2.44 ± 0.18 a	3045 ± 226 a
Dew	W1	23.43 ± 2.67 a	53 ± 6 b	2.05 ± 0.19 bc	2928 ± 247 a
	W4	17.60 ± 2.61 b	47 ± 6 b	1.76 ± 0.19 c	2826 ± 266 a
	W8	15.63 ± 2.47 b	59 ± 7 ab	1.98 ± 0.15 bc	3014 ± 235 a
Water	W1	15.42 ± 1.87 b	70 ± 7 a	2.30 ± 0.18 ab	3228 ± 240 a
	W4	19.50 ± 1.96 ab	51 ± 6 b	1.84 ± 0.15 c	2998 ± 252 a
	W8	17.71 ± 2.26 b	48 ± 5 b	1.77 ± 0.16 c	2896 ± 247 a
p-value		<0.001	<0.001	<0.001	0.40

The letters a to c correspond to the comparison of the impact of the method coupled to the duration in the same column according to an ANOVA analysis.

In the initial week of dew-retting, the fiber linear density and the proportion of coarse fiber (exceeding 34 dtex) increased, as indicated in Figure 4. As a result, there was a decrease in both tenacity and strain. These changes in linear density and mechanical properties were associated with chemical alterations in the stem and fiber. In contrast, during the first water-retting, the fiber linear density and mechanical properties stayed unchanged. However, the percentage of fine fiber (below 10 dtex) rose from 14% (W0) to 26% (W1), as shown in Figure 4. This increase in the proportion of fine fibers matched a rise in cellulose rate, Table 4 (Zimniewska 2022). So, in the initial week, the quality of water-retting fibers surpassed that of the dew-retting fibers, because the fiber was thinner and the mechanical properties were higher except the Young’s modulus.

Between W1 and W4 of dew-retting, the mechanical properties did not change, while the linear density reduced significantly from (23.43 ± 2.63) dtex to (17.60 ± 2.61) dtex. Additionally, the proportion of fine fiber (below 16 dtex) increased, similar to the rate of cellulose and hemicellulose in the bundle fiber, detailed in Table 4. The refinement of the fiber corresponded to the deterioration of the chemical elements in the stem and changes in the fiber color. From W1 to W4 of water-retting, there was no alteration in the linear density, but there was a decrease in tenacity and strain, according to Table 6. This decline in fiber quality was linked to the degradation of the chemical composition of the stem and the increase in the masses of Copper and Zinc and fiber brightness (Tables 3 and 5). By W4, both the linear density and the mechanical properties were comparable across different retting methods. This finding aligns with the research conducted by Różańska et al. (2023), though their reported linear density is lower than what has been achieved in the current study.

At the end, for both retting methods, the quality remained consistent, similar to the stem and fiber chemical composition. According to the ANOVA analysis, there was no significant difference in linear density between the methods, recorded at (14.70 ± 1.65) dtex for dew-retting and (17.71 ± 2.26) dtex for water-retting. However, this finding contrasts with Morrison et al.’s (2000) research on flax retting, which identified a significant distinction between the effects of dew- and water-retting on the fiber quality.

To conclude, from the first to the last dew-retting weeks, the fiber linear density decreased. However, the mechanical properties only decreased during the initial week, in line with results of Liu et al. (2015a). This finding is in stark contrast to Mazian et al.’s (2018) research, which indicated increases in stress, strain and Young’s modulus at the early stages of dew-retting. Unlike dew-retting, water-retting did not affect the linear density of the fiber, yet there was a notable decline in mechanical properties between the first and the final week. Ruan et al. (2015) observed in their study on flax water-retting fiber that the decrease in strain and the increase in stress during the first 10 days. For both methods, the decline of the mechanical properties corresponded to the alterations in the stem chemical composition and the fiber color. Given the results regarding the fibers, they are suitable for both the textile industry (Vandepitte et al. 2020; Zimniewska 2022) and the composite industry (Placet 2009; Réquilé et al. 2021).

Conclusion

Retting can be performed in Eastern France, in particular dew-retting, which bears similarities to practices in Western France. The duration of the retting process impacts the chemical composition and color of the stem and fiber, along with morphological and mechanical properties of the fiber, irrespective of the retting method used.

Dew-retting is recognized for its being simple and cost-effective. Conversely, water-retting has a downside of water pollution and requires an extra dying phase post-retting. According to the different analyses, water-retting is quicker than dew-retting. Therefore, the optimal duration depends on the retting method. The ideal period for water-retting is 1 week, whereas the optimal duration for dew-retting amounts to 8 weeks (which may vary depending on summer rainfall), according to assessments of the fiber’s chemical composition, morphological and mechanical properties. Given that, water facilitates the degradation of the stems and supports the growth of microorganisms. In this way, for optimizing dew-retting, mowing should be done just before the rainy season. In Eastern France, the rainy season typically begins in mid-August and extends to early September. Consequently, mowing too early in summer can extend the duration of the dew-retting process.

After 8 weeks, the stem chemical compositions are very different with a high rate of Copper and Zinc in stems using the water-retting stems and a high rate of Calcium, Manganese and Potassium in stems using the dew-retting stems. On the contrary, the fiber chemical composition remains largely comparable. Pectin serves as an effective indicator for retting. When the pectin rate is high, this suggests that the fiber is under-retted, while a low pectin rate signals over-retting. Nonetheless, this pectin rate can alter throughout the textile processing stages, influenced by mechanical actions. The stem color varies significantly at the retting end. For the textile industry, the fiber color must be homogeneous and light for the finishing process. In this case, the dew-retting fibers are too dark and blue, making the water-retting fiber color more favorable for the textile industry. Nonetheless, by the end of the retting process, the morphological and mechanical properties become similar. For the textile industry, the fiber properties must also be homogeneous too. The water-retting method is fast, making it time-efficient for the textile industry. However, this method is polluting and difficult to implement. Through mechanical actions, the textile industry process alters the morphology and mechanical properties of the fibers. As a result, these properties are meant to change during the textile process. Even though fibers retted by the water method better fit the textile industry’s needs, the dew-retting method is anyway favored due to its economic and environmental benefits.

Confirmation of authors

The authors confirm that the work described has not been published before and is not under consideration for publication anywhere else. We can also say that the publication has been approved by all of us.

Highlights

• Dew- and water-retting methods are compared over 8 weeks in Eastern France.

• For both methods, the stem chemical composition changes during the retting.

• The fiber mechanical properties change during the retting, whatever the method.

• The optimal retting duration depends on the retting method.

• The water-retting was optimized and adapted for the textile industry.

Acknowledgment

This experimentation was carried out thanks to the financing of the European Union LEADER and the Région Grand-Est. The first author acknowledges the ANR CIFRE for financing her research work and PhD.

Disclosure statement

No potential conflict of interest was reported by the author(s).