Collection and Preparation of Cocoon Shells and Fibers

Dead cocoon nests (after moth emergence) of A. panda were collected from the Kakamega rainforest, western Kenya, in May 2011. Bombyx mori cocoons (race ICIPE II) were obtained from the Commercial Insect Program of the International Centre of Insect Physiology and Ecology (icipe). Cocoons were dissected with a sharp blade to remove dead pupae and other insect exuvia from the inner sections. Leaves, small twigs, and other foreign matters were also removed from the outer surfaces of A. panda cocoons before degumming. Cocoons were soaked in hot water for 10 minutes to remove any remaining dust particles and then boiled in an aqueous solution of 5 g/L of sodium carbonate (locally known as magadi soda) for 5 hours and 1 hour for A. panda and B. mori, respectively. Degummed fibers were washed with a detergent solution comprised of 50 mL/L distilled water for 3 minutes, and 200 mL of distilled water twice. Cocoons were enclosed in wire mesh cages with a volume of 717 cm³ to ensure homogeneous boiling. Fibers were then immersed in 70% ethanol for 3 days to remove fatty and waxy substances from the surface of fibers. Cross-sectional slices of fibers were produced by pulling the fibers through an empty plastic tube and a clean cut was made using a new razor blade.

Randomly selected cocoon nests were pressed out into discs at the middle section with a sharp hole-punch for surface observation.

Scanning Electron Microscopy

For scanning electron microscopy (SEM) observation, samples were mounted onto copper stubs using double-sided tape and sputter-coated with gold for 3 minutes. The surface and cross-section of the samples were taken for SEM (Neoscape, JCM-5000; JEOL, Tokyo, Japan) under an accelerating voltage of 10 kV and a beam current of 0.1 nA.

Amino Acid Analysis

Approximately 1 mg of degummed silk was lyophilized and crushed in liquid nitrogen and then dried in an oven to remove moisture at 105°C before acid hydrolysis. Two milliliters of 6 N hydrochloric acid containing 1.0% of phenol was added to the dry silk in a vial, which was sealed under nitrogen and heated in an oven at 110°C for 24 hours. The hydrolysate was then centrifuged for 5 minutes and filtered through glass wool. The supernatant was diluted 20 times with 50% methanol before analysis. Multiple analyses were carried out using high-performance liquid chromatography (HPLC) (Prominence LGE-UV Low Pressure Gradient HPLC system; Shimadzu, Kyoto, Japan; Column Gemini C4, Length 250 mm, ID 4.6 mm, run time 25 minutes) and the resulting peaks were identified by running authentic standards of the amino acids. Cocoons of B. mori were analyzed similarly for comparison.

Determination of Weight Loss

Twenty grams of cleaned cocoon shells of A. panda and B. mori were used. The quantity of removed sericin and other cocoon components that resulted due to degumming was obtained using following equation:¹²

Determination of Moisture Regain

Moisture regain was determined using the oven drying procedure. Cocoon shells and degummed fibers of A. panda and B. mori were placed in an oven at 110°C for 24 hours. Twenty grams of oven-dried shells and fibers were then placed on a laboratory bench for 72 hours under room conditions (23°C ± 2°C and 71% ± 3% Relative Humidity (RH)). The moisture regain percentage for the cocoon shells and degummed fibers was obtained using the following equation:¹³

Tensile Testing of Fibers

Single degummed fibers (brin) of A. panda and B. mori were mounted and glued onto a 10 mm slotted rectangular cardboard, which was then fixed onto a tensile testing machine (55425 N load cell; Instron, Norwood, MA, USA). Tests were conducted with a gauge length of 10 mm at a rate of 0.1 mm/seconds at room temperature. Ten tests were conducted to generate the average stress–strain data. Cross-sectional areas of the fibers were calculated from the digital images of transverse sections on SEM micrographs and analyzed with ImageJ 1.42 q software ( http://imagej.nih.gov/). Normalized cross-sectional areas were obtained by averaging 50 brins for each sample. The tensile parameters were calculated using a designed program designed in Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) by our lab (Tensile Import v.2.0). Stress–strain curves were plotted using Origin Pro 8 (OriginLab Corporation, Northampton, MA, USA).

Thermogravimetric Analysis

Cocoon nests and degummed fiber discs (7 mm in diameter) of A. panda and B. mori were pressed out of the middle section of the cocoon shells and degummed floss, respectively, with a sharp hole-punch. Thermogravimetric analysis was performed on a TA Instruments model Q500 (TA Instruments, New Castle, DE, USA). Three tests were conducted for each sample. Temperature ranges of 25°C–900°C and 25°C–800°C were used for cocoon shells and degummed fibers, respectively, at a heating rate of 20°C/minute, as well as an N₂ flow rate of 60 mL/minute, and air cool time of 40 minutes.

Data Analysis

Percentage data were subjected to logarithmic transformation procedures to stabilize the variance before the final statistical analysis. Means were analyzed using one-way analysis of variance with a PROC analysis of variance procedure. Least significant difference tests (α = 0.05) were used to separate means.

Structures of Cocoon Nests and Degummed Fibers

Anaphe cocoon nests are sessile and fixed to branches or twigs of a tree along their side by a silk band, and they can be regarded as a large single cocoon consisting of three envelopes of silk that enclose and protect the pupae. These nests are complex cocoon structures that can consist of 20–200 individual cocoon clusters.⁵ The outer envelope is composed of very fine, but fairly strong and long fibers (Fig. 1A). Cross-sections of the outer cocoon shell layer were loosely packed with irregularly arranged fibers (Fig. 1B) and had hair-like bristles (34.29 µm in diameter) covering its surface, which is incorporated from larval spines during spinning. The bristles may be used for protecting the larvae from natural enemies such as birds (Fig. 1C). These bristles have tiny spines (diameter of 7.33 µm) arranged alternatively over their surface. These spines along with other chemicals and dusts may make working with Anaphe cocoons uncomfortable and cause allergies in workers. The middle layer consists of several closely packed sheets of silk. The innermost layer is compact and extremely tough parchment, which is made up of long and fine fibers that are embedded with high sericin gum (Fig. 1D). This provides protection for the cluster of individual cocoons located inside, and makes up to 21% of the total weight of the cocoon. The toughness of the inner layer and compacted clusters of individual cocoons may play a role in reducing the overall incidence of cocoon parasitism.^14,15 Such complicated communal nests may provide a high degree of protection in both the fire-prone regions and in the high rainfall forests, such as the Kakamega, with lower temperatures and long rainy seasons, which would be an evolutionary advantage. Generally, fibers were smooth with few fractures and had bamboo-like structures with nodes at certain intervals (Fig. 1E). The cross-sectional shape of fibers varied from globular to curved (crescent) (Fig. 1F). Double trapdoors (tunnels) constructed by the larvae on the cocoon nests and individual cocoons allow the moth to pass through after emergence. The tunnels may also be used as air exchange windows that condition the inside environment for the larvae and pupae during spinning and dormancy.

Figure 1.

SEM micrographs of Anaphe panda cocoon shell and fiber surfaces and cross-sections. (A) Surface of outer layer; (B) cross-section of outer cocoon layer; (C) hairs on cocoon surface; (D) cross-section of inner cocoon layer; (E) nodes on the fiber outer surface; (F) cross-section.

Amino Acid Composition of Fibroin

The results showed that both A. panda and B. mori were characterized by the presence of high proportions of the three amino acids, glycine, alanine, and serine (Table 1). The three amino acids represented approximately 80% and 89% of the total amino acids present in A. panda and B. mori, respectively. Alanine was the dominant amino acid (34%) followed by glycine and serine in A. panda; however, in B. mori glycine (43%) was predominant. Other amino acids, including glutamic acid, asparatic acid, valine, and methionine were also found in various amounts. The proportions reported here were comparable with those reported for eri, tasar, and muga (34%–36%).¹⁶ However, the alanine content of A. panda was lower than other nonmulberry silk fibroins, particularly for muga (44%).¹⁷ Alanine content as high as 52%–59% has also been reported for the outer layer of A. reticulate and the sum of alanine and glycine content accounted for more than 90%.^11,18 The composition of some of the amino acids such as tryptophan, cystine, asparagine, glutamine, and isoluecine could not be included because their analysis and degradation products were inhibited by their instability in oxidative or strongly acidic media, deamination, slow cleavage, and/or light.

Table 1.

Composition of major amino acids in the fibroin of Anaphe panda and Bombyx mori (ratios mol/g).

Weight Loss and Moisture Regain of Cocoon Shells and Degummed Fibers

Table 2 shows the moisture regain and weight loss of the cocoon shells and degummed fibers of A. panda. The total weight loss due to degumming was 25.6%; however, the traces of sericin gum observed on the fibers (Fig. 1B) suggested that the degumming process did not completely remove all of the gum present on the fibers. The weight loss reported here was higher than for other wild silks such as Antheraea proylei, A. assama, A. pernyi, and A. yamamai (7%–13%).⁹ However, A. panda had less sericin gum than Gonometa postica, which showed more than 50% weight loss.¹⁹ The nature and chemical composition of sericin protein considerably affects its solubility in hot water and contributes to the amount of total weight loss from silk cocoon shells. Degummed fibers showed higher moisture regain (9.9%) than cocoon nests (8%). This may be due to the presence of hairs and other foreign matter on the surface of the cocoon nest, which deny access to moisture absorption. The moisture regain of the A. panda cocoon nest and fibers were higher than for B. mori, as reported by Sen and Babu.¹⁶ They also reported that moisture regain of nonmulberry silks is higher than that of their domestic counterparts. This may be due to the difference in compactness, as well as the change in quantity and composition of the sericin and fibroin (hydrophilic/hydrophobic amino acids) in the cocoon shells and degummed fibers. A. panda contained a higher hydrophilic/hydrophobic amino acid ratio (0.23) than B. mori (0.19) (Table 1), which may have contributed to the higher moisture regain.

Table 2.

Weight loss and moisture regain of Anaphe panda and Bombyx mori silk fibers and cocoon shells.

Tensile and Thermal Degradation Properties

Figure 2 shows the stress–strain curves of single A. panda and B. mori fibers. The curves were widely scattered for A. panda. Such variation in mechanical properties is inevitable for wild silk fibers due to their irregular cross-sectional area throughout the fiber axis and across the different layers of the cocoon nest. The stress–strain curves of both A. panda and B. mori showed two regions: an initial elastic region directly, followed by a strain-hardening region. A. panda fibers showed a lower breaking strain (15.4%) and comparable breaking stress (0.4 GPa) as B. mori (Table 3). In contrast, Sen and Babu¹⁶ reported higher breaking extensions for nonmulberry silk fibers. Gheysens et al¹⁹ also reported higher (37%) breaking strains for G. postica. The extended time of boiling for A. panda in the alkaline solution coupled with a possible difference in the proportion of the amorphous and crystalline region of the fiber may have contributed to the difference in the tensile properties observed.

Figure 2.

Stress–stress curves of degummed single fibers of (A) Anaphe panda and (B) Bombyx mori.

Table 3.

Tensile properties and thermogravimetric weight loss of Anaphe panda and Bombyx mori degummed fibers

Figure 3 shows the thermogravimetric curves of A. panda and B. mori degummed fibers and cocoon nests. Initial rapid weight loss occurred due to the loss of water from the cocoon shells and fibers followed by a period of steadiness. However, with increased temperature, a rapid decline in weight followed due to decomposition. This occurred at 314°C and 293°C for A. panda and B. mori cocoon shells, respectively. B. mori fibers and cocoon shells had the highest total weight loss (94.8% and 93.5%, respectively) (Table 3). However, A. panda cocoon shells showed better heat resistance with both dehydration and decomposition occurring at higher temperatures than B. mori. The higher heat stability in the cocoon shells, as compared to the fibers, may indicate a role in thermoregulation. Danks²⁰ also reported the role that cocoon shells play in the acquisition and conservation of heat. Both cocoon shells and degummed fibers underwent a multistage weight loss scenario, which may be due to the difference in the polymorphs of crystalline structures, and due to the amino acid composition.²⁰ Zhang et al²² also reported similar decomposition behavior for B. mori cocoon shells, as illustrated by rapid weight losses occurring at 98°C and 280°C.

Figure 3.

TGA curves of Anaphe panda fibers and cocoon shells.