PHYSICAL REVIEW E 88, 062605 (2013) Creep anomaly in electrospun fibers made of globular proteins Omri Regev, 1 Arkadii Arinstein, 2 and Eyal Zussman 1, 2 , * 1 Russell Berrie Nanotechnology Institute, Technion–Israel Institute of Technology, Haifa 32000, Israel 2 Department of Mechanical Engineering, Technion–Israel Institute of Technology, Haifa 32000, Israel (Received 11 October 2012; revised manuscript received 21 October 2013; published 30 December 2013) The anomalous responses of electrospun nanofibers and film fabricated of unfolded bovine serum albumin (BSA) under constant stress (creep) is observed. In contrast to typical creep behavior of viscoelastic materials demonstrating (after immediate elastic response) a time-dependent elongation, in case of low applied stresses (<1 MPa) the immediate elastic response of BSA samples is followed by gradual contraction up to 2%. Under higher stresses (2–6 MPa) the contraction phase changes into elongation; and in case of stresses above 7 MPa only elongation was observed, with no initial contraction. The anomalous creep behavior was not observed when the BSA samples were subjected to additional creep cycles independently on the stress level. The above anomaly, which was not observed before either for viscoelastic solids or for polymers, is related to specific protein features, namely, to the ability to fold. We hypothesize that the phenomenon is caused by folding of BSA macromolecules into dry molten globule states, feasible after cross-linked bonds break up, resulting from the applied external force. DOI: 10.1103/PhysRevE.88.062605 PACS number(s): 36.20.−r, 87.15.La, 62.20.Hg I. INTRODUCTION When an ideal elastic material is subjected to a con- stant stress, its immediate strain response remains constant. In contrast, viscoelastic polymers undergo creep, a time- dependent viscoplastic response, in which molecular motion (reptation) results in monotonic elongation of the material over time [1–3]. A material’s creep response can be dictated by specific structural features (e.g., degree of polymer matrix heterogeneity) and presents a great challenge in engineering applications, where designers try to increase the creep resis- tance of polymers [4,5] and composite materials [6–9]. In biopolymers [10], biofibers [11], and soft biological tissues [12], creep response may indicate structural and biochemical changes [13]. Herein, we report a creep anomaly observed in structures made of bovine serum albumin (BSA). The BSA was de- natured, which involved reduction of disulfide bonds and concomitant controlled protein unfolding, providing for the reformation of new, extended, polymer-like structures rich in strong inter- and intramolecular disulfide bonds [14]. When exposed to tensile stress, the denatured material demonstrated a nonmonotonic creep response, with an initial rapid elastic elongation phase, followed by an unexpected contraction (decreased creep rate), and only thereafter a phase which exhibited creep behavior typical of polymers (gradual increase of creep rate). A similar nonmonotonic time response was reported in creep experiments of ultrahigh-molecular-weight polyethylene [15]; however, unlike the BSA, the contraction was an order of magnitude lower and was instead related to relaxation after intensive preloading. II. MATERIALS AND METHODS Experiments were conducted on BSA fibers electro- spun [14,16] from a solution of 10 wt% BSA in 2,2,2- trifluoroethanol (TFE), distilled water, and 2-mercaptoethanol * meeyal@technion.ac.il (ME; 0.2 g per 1 g BSA). Aligned fiber mats were collected on a vertical rotating wheel (1500 rpm) [17]. Films of BSA were prepared by solution casting on Teflon dishes. All samples were stored for at least one month in a vacuum of ∼10 −3 atm, and then stored at room temperature under ambient conditions for 2 or 18 months after fabrication. A dynamic mechanical analyzer (DMA) was used to test rectangular-shaped samples (10 × 5 × 0.1 mm 3 ) mounted using a torque-meter (2 lb-in.). Load was applied along the fiber axis of the mat. All tests were conducted at room temperature, in 40–70% relative humidity. III. RESULTS Quasistatic tensile tests (strain rate of 1%/min) showed that fiber mat strength was 13.0 ± 1.6 MPa (n = 12), with 5.8 ± 1.2% extensibility, and an elastic modulus of 514 ± 121 MPa (calculated over a strain range of 0.2–0.8%). Figure 1 shows the results of creep experiments conducted under tensile stresses of 1, 4, and 8 MPa. Fiber mats broke up above stresses of ∼9 MPa. At stresses below 5 MPa, fiber mats demonstrated a unique contraction behavior (see Fig. 1). The creep response started with an elastic response that lasted 0.36 s, followed by continuous elongation for 0.5–20 s and gradual contraction over time. At low stresses (<5 MPa) contraction was dominant, while contraction was suppressed by creep elongation when stresses have exceeded 6 MPa. When a mounted fiber mat was tested in the second cycle the contraction creep phase was absent, while the creep rate slowly increased with time, as expected of polymeric viscoelastic materials (see Fig. 2). This response was observed for all tested stresses between 1 to 8 MPa. The elastic modulus of fiber mats in the second creep cycle was higher in 350 ± 100 MPa than that of the first cycle (see the instantaneous response in Fig. 2), a change which proved to be stress independent. At first glance, the creep anomaly of BSA fiber mats may be attributed to the strong extension of the polymer matrix during the electrospinning process [19] and, possibly, to confinement [20]. However, the unexpected contraction was also observed in films cast from the same BSA solution used for fiber fabrication (see Fig. 3). In both cases, the BSA chains are 1539-3755/2013/88(6)/062605(5) 062605-1 ©2013 American Physical Society