Fabrication of 3D Nanofiber Scaffold by Sequential Electrospinning Stephanie Polanco Martínez Master of Engineering in Manufacturing Engineering Advisor: José A. Morales, Ph.D. Industrial Engineering Department Polytechnic University of Puerto Rico Abstract  The generation of a 3D porous scaffold morphologies of the ECM vary according to the that would mimic the native ECM of tissue and function of the tissue, and even though some ECM organs is a topic of interest in tissue engineering. have 2D characteristics (eg. top layer of skin), Electrospinning is a widely used technique in the fibrous structures with 3D orientation can be found area because it generates fibers of nano metric in breast, liver, bladder and lung tissues, among diameter with high surface, so several studies have others [3]. 3D nanostructures with a significant applied it in the creation of 3D scaffolds. This thickness are increasingly being investigated for research proposes a new approach that consists on tissue engineering. the application of sequential electrospinning cycles Electrospinning is a process in which polymer over a rotational collector leaving a timespan fibers are formed by the application of a strong between each cycle in which all the remaining electric field on a polymer solution or melt. This solvent evaporates and the electro charges on the technique has raised great interests among tissue upper layer are dissipated. It was proved that the engineers because electro spun nanofibers present thickness of the scaffold increases linearly with extremely high surface area, and thus work as an each cycle. Additionally, when compared to excellent platform to study the cell-environment continuous electrospinning the novel method interactions, effects of growth factors, drug generates a scaffold which presents almost double delivery, etc., [4] [5]. One of the key requirements the thickness size and a similar porosity. of scaffolds is to replicate the topographies and Key Terms  3D Scaffold, Porous spatial structures of native ECM in order to Architecture, Sequential Electospinning, Tissue facilitate cellular penetration, growth and Engineering. differentiation [3] [6]. The 3D pore structure is one of the critical features for an ideal scaffold [7]. INTRODUCTION However, the electrospinning process has generated mostly 2D thin layers (usually between 200 to 1000 Tissue engineering is an emerging nm [8] of compact nanofibers with small pores thus multidisciplinary field that focuses on design and limiting its application in tissue engineering [9]. development of functional, three-dimensional human tissue for medical research and therapeutic RESEARCH DESCRIPTION applications. One of the key research areas is to develop a scaffold that mimics the extracellular Traditional electrospinning presents difficulties matrix to promote tissue regeneration. This is a to generate 3D scaffolds, in the first place, due to complex arrangement of biomaterials in micro to the repulsion generated by the accumulation of nano scale that form a supportive structure for cells, electrostatic charge between the already deposited promoting their growth, adhesion, migration and fibers and the incoming ones [8], and secondly, differentiation [1]. Mimicking the 3D porous because the fibers that reach the collector are wet architecture of the ECM is essential for the success and soft (the solvent doesn’t evaporate completely of the scaffold, the pore network must be suitable during the flight) making the 3D configuration of for cellular ingrowth, promote cell to cell contact the jet collapse into a 2D layer when deposited on [1], and allow the diffusion of nutrients, metabolites the collector [10]. Several studies have addressed and soluble factors [2]. It must be considered that this issue by testing new techniques such as wet electrospinning, coarse fibers integrated electrospinning method. The travel distance used electrospinning, and electrospinning with porogens for this setup was 63 mm and the duration of the [6]. electrospinning was 2 hours. Notice the ratio travel To address this problem, this study will test if a distance vs. process time is almost the same for thick and porous scaffold can be developed by both setups. using the sequential electrospinning technique. This Material work presents a novel facile electrospinning method for creating 3D nanofiber scaffolds. The Poly vinyl alcohol (10wt%, PVA, Tongli Tech, scaffolds were fabricated by sequentially China) water solution was used as the nanofiber electrospinning the polymer nanofibers on a material. Fabrication of 3D Nanofiber Scaffold by solidified layer. Between each electrospinning Sequential Electrospinning. [11] cycle, the scaffolds were left to dry in order to Electrospinning Setup avoid fiber dissolution during continuous A stainless steel cylinder collector with a electrospinning. diameter = 100 mm, length = 300 mm was installed inside the Electrospinning Robotic Platform (TL- Pro-BM, Tongli Tech, China). The collector was located perpendicular to the spinneret to create a lineal fiber jet travel distance; this collector was connected to a negative voltage and rotated at speed of 1000 rpm. A stainless-steel needle was assembled on the spinneret with a 23 gauge (0.33 mm inner diameter) which is connected to positive voltage supplier. The tip of syringe had a fixed distance of 122 mm to the collector with a fixed Figure 1 angle of 45°. It was moved parallel to the collector Electrospinning with a Cylinder Collector at speed of 5 mm/s. The configuration of the machine is presented in Figure 1, and the METHODOLOGY parameters described in Table 1. An eight-layer scaffold was built using Table 1 Fixed Parameter for the Electrospinning Configuration sequential electrospinning technique, with a start point of 10 mm and a travel distance of 250 mm. TTD PV NV MPR RS SSS (mm) (kv) (kv) (ml/h) (rpm) (mm) Each layer was produced in one hour long electrospinning cycles maintaining the parameters 122 10 10 2 1000 5 indicated in Table 1. To let the layers properly dry and the dispersion of electro charges, there was at Attributes Measurement least a 3 hours long pause in between the cycles. The attributes of interest for this study are Before starting a new cycle one small sample (1 cm thickness, diameter size, porosity, and orientation, x 5 cm app) of mat was peeled off the collector. to compare the actual process calling continuous The cycle was repeated 8 times, which mean 8 electrospinning with the proposal one, called hours of electrospinning were completed. Each sequential electrospinning. The focus in this sample was named after the amount of layers it research will be the thickness and porosity. For includes, e.g. to generate sample 3 three cycles of each sample the thickness was measure with a sequential electrospinning where completed. Also, thickness gauge (Mitutoyo, Japan) and direct with a control mat was built using the traditional the SEM images. Images of samples 1, 8 and electrospinning cycle is adding approximately control were obtained with the scanning electron 0.0045 μm to the final scaffold. microscope (Hitachi S4300SE/N). The magnifications used were 700k and 1500k for plane view, and 800k for cross sectional view. The images were then analyzed to obtain the required attributes with ImageJ open source software. The thickness of the scaffold was measure manually from the image. Porosity, orientation and diameter size were calculated by the Diameter J and Porosity J plugin program. Figure 3 Sample Thickness Measured by Gauge ANALYSIS & RESULTS The thickness of sample 1, 8 and control was also measured using the SEM cross sectional Firstly, the electro spun samples were visually images presented in Figure 4. The difference in examined to distinguish observable attributes such thickness is observable in the photographs. To have as color or hardness. Photo A is the layer 1 with a better precision of the thickness it’s was measure sequential electrospinning, photo B is a new the five different parts in the layer to calculate the layer over dry electro spun and this create layer average for each one. These values are compared to two. those obtained with the gauge. It can be observed that using the latter tool results in consistently smaller thickness values compared to those by SEM. Figure 2 Figure 4 Eight-Layer Sample after the Process of Sequential Cross section of Mat Samples Electrospinning A. Sample 1. B. Sample 8. C. Control sample. Thickness Diameter and Porosity The thickness values of the samples obtained The images obtained with the SEM are with the gauge are presented as a scatter plot presented in Figure 5. It can be observed that the adjusted to a linear curve (Figure 3). It is clear that control sample has the smallest fiber diameters and the points follow a linear tendency with an R2 the fibers have no orientation. Sample 4, on the value of 98%. Each additional sequential other hand, has larger pores. In order to validate this observations, porosity, fiber diameter and orientation were measured in the images 1500x of magnification by the program plugins Diameter J and Orientation J. Figure 7 Diagram of the Fiber Diameter Fiber Orientation Orientation J software generates a histogram of the orientation angle of the fibers for each image analyzed. These images are presented in Figures 8, 9 and 10 for sample 1, 8 and control, respectively. The orientation histogram for sample 1 has a concave up shape, with most fibers having a -90° or 90° orientation. This means that fibers are mostly Figure 5 oriented vertically. For sample 8, the tendency can Scanning Electron Microscopy (SEM) Images Lines: A. Sample 1. B. Sample 8. C. Control sample be detected but it’s not as clear as with sample 1. Columns: 1. 700x. 2. 1,500x. On the other hand, the histograms for the control sample are completely random therefore the fibers Porosity percentages are presented in Figure 6. have no constant orientation, demonstrating the The most porous sample is control sample, observation made before with the SEM image. followed by sample 8. Figure 8 Fiber Orientation of Sample 1 Two Images taken with SEM x1500 Figure 6 Diagram of the Porosity Percentage Figure 9 Fiber Orientation of Sample 8 The porosity values don’t vary widely between Two Images taken with SEM x1500 each other, with the proposal as being 47% and control as 51%, the error marge between then is 7%. The fiber diameter values are displayed in Figure 7. In this case there’s more difference between the samples, as Sample 1 has almost Figure 10 double fiber diameter than sample control as had Fiber Orientation of Control Sample stipulate that it would be the expect value. Two Images taken with SEM x1500 Statistical Analysis thickness added to scaffold behaves in a lineal way, according to this results, the desire thickness can be One way Anova was performed in Minitab to obtained by repetition of cycles. However, it is compare the different data sets and significance was encouraged to realize the experiment with an accepted when the p-value was less than 0.05. As additional number of cycles to prove that the the result, the P value of the hypothesis was thickness curve maintains its linearity. Another computed and the p-value for the coefficient is less drawback of using the technique proposed is the than 5%, close to 0, thus it is significant impact in time it takes to fabricate the scaffold, considering the results. that before starting with the new layer the last one Also the average and standard deviation was has to be dried. If each cycle only adds 10 microns calculate for each attribute to know how much the of thickness then to be able to reach 1 mm around a values can moved away from the average and the 100 cycles must be completed. In addition to the probability that the data move up or down polymer used that has to be ordered from China and according to the general process. The data does not the shipment takes a time. deviate to much between them, this conclusion is very important because it is a medicinal process DISCUSSION and the closer the data are between them, the less error will be processed in order to obtain a better Several studies in the field of tissue homogeneity. engineering have focused on the construction of a The proposed margin of error is 5%, this will scaffold with a 3D pore architecture utilizing the help us to understand that the data is not precise or electrospinning technique. One of the main exact, we don’t choose a grater margin of error approaches consists in the variation of the collector because we are working on human observations and plate from a 2D to 3D shape, which can vary in we need precision data, also to have a smaller complexity from parallel plates to other custom- percentage of error the sample needs to be larger. built configuration [12]. Additional studies have Also with a confidence level of 95% we get a Z of used a liquid collector. Its working principle is 1.96, with this number we calculate the sample size. immersion precipitation, fibers solidified instantly The sample size helps us to understand better how as they are deposited in a non-solvent liquid [13]. much quantity of sample we need to collect to Another method used to obtain thicker layers is avoid the bias in the interpretation of the data, as it post process of the traditional scaffold, either by helps us reduce costs and time. The greater sample stacking layers together [14] or by folding them size we get was for the continuous electrospinning into a tubular shape [15] [16]. To improve the (traditional one) for the diameter size with an N of porosity of the scaffolds porogen particles and 47, this tells us that to create a better scaffold micrometer-sized fibers have been introduced diameter should be done 47 samples which during the electrospinning process thus increasing indicates that it would be 47 hours of continuous the void between the nanofibers. These approaches work. But with the sequential electrospinning for have had positive results, however, they add the sample of diameter only need a sample size of materials and steps to the traditional 27. electrospinning process, and some may have difficulties with the preservation of the mat Limitation morphology during the removal from the collector The total thickness of the mat built by [17]. sequential electrospinning reached only the tens of Sequential electrospinning, as proposed in this microns, when in tissue engineering a functioning paper, consists on the successive application of scaffold must have a 3D structure with at least a electrospinning over dried electrospun layers thus millimetric thickness. It was discovered that the creating a thicker layer with each cycle to generate improve the porosity and increase the thickness 3D porous scaffolds. It varies from traditional achieved. electrospinning as the latter is a continuous process, while in the former, a new electrospinning cycle is RECOMMENDATION only started when the last layer deposited has Based on the results, the final recommended is already dried. With the study it was verified that the a study focused in determining what is the thickness of the scaffold increases linearly with maximum amount of material applied in each cycle each sequential cycle. Additionally, it was is proposed [18]. Although the porosity of the mat confirmed that realizing the traditional method of wasn’t affected with the repeating cycles, as stated electrospinning, maintaining the same parameters in the objected, it is still not enough for tissue and time-travel distance proportion, generates a engineering application. To correct this, the study thinner scaffold. For this specific study the scaffold could be repeated using two polymer solutions built by sequential electrospinning reached almost which dissolve in different solvents. After building double the thickness and increases linearly with the electro spun scaffold one of the polymers is each cycle of the one by continuous dissolved, leaving a larger pore size. Different electrospinning. It was expected to obtain a thinner pump rates should be experimented with to obtain and densely packed continuous electro spun different porosities. It would be interesting to scaffold. However, the control sample presented a continue further studies to improve the porosity and higher porosity than the sample obtained by increase the thickness achieved [19]. sequential electrospinning. This indicates that the control sample has less amount of polymer REFERENCES deposited. One possible explanation of this phenomenon is that the accumulation of [1] B. Wulkersdorfer, K. K. Kao, V. G. Agopian, A. Ahn, J. C. Dunn, B. M. Wu & M. Stelzner, “Bimodal Porous electrocharge on the collector, as the Scaffolds by Sequential Electrospinning of Poly(glycolic electrospinning process continues, weakens the acid) with Sucrose Particles,” in International Journal of charge density between needle and collector. Polymer Science, 2010. [2] J. G. Kim & W. Kim, “Highly Porous 3D Nanofiber CONCLUSIONS Scaffold Using an Electrospinning Technique,” in Journal of Biomedical Materials Research Part B: Applied A novel approach for electrospinning was Biomaterials, pp. 104-110, 2006. presented to generate 3D porous scaffolds. [3] S. Cai, H. Xu, Q. Jiang & Y. Yang, “Novel 3D Electrospun Sequential electrospinning consists of the repetition Scaffolds with Fibers Oriented Randomly and Evenly in of electrospinning cycles on the same collector Three Dimensions to Closely Mimic the Unique separated by a drying time gap. It was proved that Architectures of Extracellular Matrices in Soft Tissues: Fabrication and Mechanism Study,” in Langmuir, vol. 29, the thickness of the scaffold increases linearly with pp. 2311-2318, 2013. each cycle. Scaffolds built with the sequential and [4] S. Tan, X. Huang & B. Wu, “Some fascinating phenomena the continuous methods were compared. The in electrospinning processes and applications of thickness obtained using the novel method almost electrospun fibers,” in Polymer International, vol. 56, pp. doubles the traditional one. Although the porosities 1330-1339, 2007. of both scaffolds are similar, the traditional method [5] D. Liang, B. Hsiao & B. Chu, “Functional electrospun achieved a bigger porosity percentage. It would be nanofibrous scaffolds for biomedical applications,” in interesting to continue further studies to improve Advance Drug Delivery Reviews, vol. 59, pp. 1392-1412, 2007. the porosity and increase the thickness achieved. It would be interesting to continue further studies to [6] F. J. O'Brian, “Biomaterials & scaffolds for tissue engineering,” in Materials Today, vol. 14, no. 3, pp. 88-95, March 2011. [7] R. Bagherzadeh, S. S. Najar, M. Latifi, M. A. Tehran & L. [19] Z. M. Huang, Y. Z. Zhang, M. Kotaki & S. Ramakrishna, Kong, “A theoretical analysis and prediction of pore size “A review on polymer nanofibers by electrospinning and and pore size distribution in electrospun multilayer their applications in nanocomposites”, in Composites nanofibrous materials,” in Society for Biomaterials, vol. Science and Technology, vol. 63, pp. 2223–2253, 2003. 101, pp. 2107-2117, 2013. [8] D. Ahirwal, A. Hébraud, R. Kádár, M. Wilhelm & G. Schlatter, “From self-assembly of electrospun nanofibers to 3D cm thick hierarchical foams,” in Soft Matter, vol. 9, pp. 3164-3172, 2013. [9] F. Hejazi, H. Mirzadeh, N. Contessi, M. C. Tanzi & S. Faré, “Novel class of collector in electrospinning device for the fabrication of 3D nanofibrous structure for large defect load-bearing tissue engineering application,” in Society for Biomaterials, vol. 105, pp. 1535-1548, 2017. [10] B. Sun, Y. Z. Long, F. Yu, M. M. Li, H. D. Zhang, W. J. Li & T. X. Xu, “Self-assembly of a three-dimensional fibrous polymer sponge by electrospinning,” in Nanoescale, vol. 4, pp. 2134-2137, 2012. [11] D. Lina, R. Leonardo & L. Marcos, “Electrospinning: La era de las nanofibras,” in Revista Ibeomericana de Polimeros, vol. 14(1), pp 1-2, 2013. [12] lsberg & S. A. Khan, “Three-Dimensional Electrospun Alginate Nanofiber Mats via Tailored Charge Repulsions,” in Small, vol. 8, no. 12, pp. 1928-1936, 2012. [13] E. Kostakova, M. Seps, P. Pokorny & D. Lukas, “Study of polycaprolactone wet electrospinning process,” in eXPRESS Polymer Letters, vol. 8, no. 8, pp. 554-564, 2014. [14] Y. Yang, I. Wimpenny & M. Ahearne, “Portable nanofiber meshes dictate cell orientation throughout three- dimensional hydrogels,” in Nanomedicine: Nanotechnology, Biology and Medicine, vol. 7, pp. 131- 136, 2011. [15] M. Deng, S. G. Kumbar, L. S. Nair, A. L. Weikel, H. R. Allcock & C. T. Laurencin, “Biomimetic Structures: Biological Implications of Dipeptide-Substituted Polyphosphazene-Polyester Blend Nanofiber Matrices for Load-Bearing Bone Regeneration,” in Advanced Functional Materials, vol. 21, pp. 2641, 2011. [16] L. D. Wright, R. T. Young, T. Andric & J. W. Freeman, “Fabrication and mechanical characterization of 3D electrospun scaffolds for tissue engineering,” in Biomedical Materials, vol. 5, no. 5, pp. 055006, 2010. [17] C. Thompson, G. Chase, A. Yarin & D. Reneker, “Effects of parameters on nanofiber diameter determined from electrospinning model,” in Polymer, vol. 48, pp. 6913- 6922, 2007. [18] H. Karakas, “Electospinning of nanofibers and their applications,” in Electrospinning, 2010.