Jump to ContentJump to Main Navigation
Show Summary Details
More options …

BioNanoMaterials

Editor-in-Chief: Jockenhoevel, Stefan / Stiesch, Meike / Barcikowski, Stephan / Sternberg, Katrin

Editorial Board Member: Muellen, Klaus / Weitschies, Werner / Liefeith, Klaus / Boccaccini, Aldo R. / Williams, J. Koudy / O'Brien, Fergal J. / Kellomäki, Minna

4 Issues per year


CiteScore 2016: 0.59

SCImago Journal Rank (SJR) 2016: 0.241
Source Normalized Impact per Paper (SNIP) 2016: 0.172

Open Access
Online
ISSN
2193-066X
See all formats and pricing
More options …
Volume 17, Issue 3-4 (Sep 2016)

Issues

Fibre pulsing during melt electrospinning writing

Gernot Hochleitner
  • Department for Functional Materials in Medicine and Dentistry, University Hospital Würzburg, Pleicherwall 2, 97070 Würzburg, Germany
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Almoatazbellah Youssef
  • Department for Functional Materials in Medicine and Dentistry, University Hospital Würzburg, Pleicherwall 2, 97070 Würzburg, Germany
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Andrei Hrynevich
  • Department for Functional Materials in Medicine and Dentistry, University Hospital Würzburg, Pleicherwall 2, 97070 Würzburg, Germany
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Jodie N. Haigh
  • Nanotechnology and Molecular Science Discipline, Science and Engineering Faculty, Queensland University of Technology, 2 George St, Brisbane 4001, Australia
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Tomasz Jungst
  • Department for Functional Materials in Medicine and Dentistry, University Hospital Würzburg, Pleicherwall 2, 97070 Würzburg, Germany
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Jürgen Groll
  • Department for Functional Materials in Medicine and Dentistry, University Hospital Würzburg, Pleicherwall 2, 97070 Würzburg, Germany
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Paul D. Dalton
  • Corresponding author
  • Department for Functional Materials in Medicine and Dentistry, University Hospital Würzburg, Pleicherwall 2, 97070 Würzburg, Germany
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
Published Online: 2016-05-19 | DOI: https://doi.org/10.1515/bnm-2015-0022

Abstract

Additive manufacturing with electrohydrodynamic direct writing is a promising approach for the production of polymeric microscale objects. In this study we investigate the stability of one such process, melt electrospinning writing, to maintain accurate placement of the deposited fibre throughout the entire print. The influence of acceleration voltage and feeding pressure on the deposited poly(ε-caprolactone) fibre homogeneity is described, and how this affects the variable lag of the jet drawn by the collector movement. Three classes of diameter instabilities were observed that led to poor printing quality: (1) temporary pulsing, (2) continuous pulsing, and (3) regular long bead defects. No breakup of the electrified jet was observed for any of the experiments. A simple approach is presented for the melt electrospinning user to evaluate fibre writing integrity, and adjust the processing parameters accordingly to achieve reproducible and constant diameter fibres.

This article offers supplementary material which is provided at the end of the article.

Keywords: additive manufacturing; electrospinning; polymer processing; scaffold; tissue engineering

Introduction

There are many polymer processing technologies available to manufacture scaffolds for tissue engineering (TE) [1], [2], [3]. Some, such as porogen leaching [4] or thermally induced phase separation [5], were initially developed decades ago, while more recent additive manufacturing (AM) principles have also been used to fabricate scaffolds [1], [6], [7]. In contrast to other processing techniques, AM can be used for superior shaping of the printed scaffolds by using a computer-aided process to design and control the porosity, mesh-width, fibre or strut diameter, and wall thickness in three dimensions [7], [8]. This processing flexibility can be used to tailor structural gradient or zonal constructs and is therefore particularly interesting for TE approaches and scaffold manufacturing [9], [10].

Recently, melt electrospinning writing (MEW) has been described as an AM technology for scaffold-based TE applications [11], [12]. MEW is based on electrohydrodynamic direct writing, where a molten jet is stabilised at very low flow rates with an accelerating voltage. This electric stabilisation of a jet [13] (i.e. with no droplet formation) at low flow rates provides a predictable jet path that allows continuous direct writing and layer by layer assembly. This results in fibres, that are substantially smaller than those produced by fused deposition modelling (FDM) [8], [14], typically 5–40 μm [12], [15], [16] and as small as 800 nm [17]. Importantly, the fibres can be accurately deposited upon each other, so that three-dimensional (3D) structures can be designed and built to millimetre thicknesses [18].

While MEW is a relatively new processing technology, there is an increasing body of evidence suggesting that there are multiple advantages to using it for fabricating scaffolds [18], [19], [20], [21], [22]. MEW is solvent-free – therefore the need to print within a fume hood or to use special equipment to remove volatile chemicals is not required, and solvent accumulation in fibres does not occur. Another benefit of MEW is that small fibre diameters lead to flexible constructs that enable even relatively rigid polymers to be fabricated as soft, compliant structures [16], [20], [21]. So far, melt electrospinning has been used to process the following medical polymers: poly(ε-caprolactone) (PCL) [12], [16], [17], [18], [21], [22], [23], poly(lactide-co-glycolide) [24], poly(lactic acid) [25], PCL-block-poly(ethylene glycol) [26], [27], [28], [29], [30], poly(lactide-co-caprolactone-co-acryloyl carbonate) [31], polypropylene [28], [32], poly(methyl methacrylate) [33], and poly(2-ethyl-2-oxazoline) (PEtOx) [15].

When comparing MEW to standard solution electrospinning, it is important to note that the polymer melts used have significantly different physical properties than a typical polymer solution. Both the fluid conductivity and viscosity vary and this influences fabrication aspects such as the final diameter [34], electrified jet path [35] and Coulomb interaction of previously deposited fibres. One result of this is that direct writing from melts can be performed at a relatively large range of distances between the spinneret and the collector, from hundreds of microns up to 10 cm [29], [36]. In case of melt electrospinning the instrumental parameters, including the electrification of the jet, must be adjusted to each other in order to generate a stable process leading to a high quality fibre deposition.

For MEW, when the mass flow rates from the spinneret onto the collector are not adjusted, the phenomenon of “pulsing” occurs. Herein, “pulsing” is defined as the undesired sectional oscillation of the fibre diameter. This is a serious polymer processing issue to overcome for MEW, since it affects the quality of the printed structure. Pulsing also influences the fibre placement when there is a change in direction of the collector stage (i.e. turning). Therefore, key scaffold characteristics such as stiffness, fibre spacing or specific surface area of the printed constructs will not be constant when fibre pulsing occurs (Figure S1). In this research paper, we will describe certain conditions where the fibre pulsing instability manifests and how to eradicate this phenomenon from MEW scaffolds by varying the instrumental parameters: acceleration voltage and feeding pressure. The samples were made of PCL, a long-term hydrolytically degradable polymer that is increasingly used in TE research [37] and the clinic [38]. Scanning electron microscopy (SEM) analysis of fibres was done using both stable and pulsing conditions – and produced grids which are the unit repeat structure for TE scaffolds made by MEW [17], [18]. Furthermore, we present a simple approach to visually detect and overcome pulsing that can be used as a guideline for manufacturing high quality TE scaffolds.

Experimental methods

Materials

PCL was sourced from Corbion Inc. (PURASORB PC 12, Lot# 1412000249, 03/2015, Gorinchem, Netherlands) and used throughout this study. To avoid polymer degradation during storage, the PCL was repacked into 50 mL Falcon tubes, under a N2 atmosphere in a glove box to ensure proper storage conditions. Once repacked, the polymer was stored at –80 °C.

MEW device

Experiments were performed with a custom-built MEW device based on a commercially available CNC platform. Within an aluminium frame box (68×60×80 cm) with polycarbonate windows (Figure 1), an x-y slide system (Bosch Rexroth AG, Lohr am Main, Germany) was mounted on an aluminium plate and connected to programmable logic controllers (Bosch Rexroth AG, Lohr am Main, Germany). An electrical heater was utilised that contained disposable syringes (Nordson EFD Deutschland GmbH, Pforzheim, Germany) filled with approximately 0.5 mL–1 mL PCL and linked to a digitally controlled air pressure system (Bosch Rexroth AG, Lohr am Main, Germany). A 22 G spinneret (Nordson EFD Deutschland GmbH, Pforzheim, Germany), with a length of 12.5 mm protruded 0.5 mm out of the high voltage (HV) electrode supplied by a positive HV source (LNC 10000-5 pos, Heinzinger Electronic GmbH, Rosenheim, Germany). The negative HV source (LNC 10000-5 neg, Heinzinger Electronic GmbH, Rosenheim, Germany) was connected to a 200×175 mm stainless steel platform (with the collector placed on this platform), attached to the x-y slide system.

Schematic of a MEW device. In order to use MEW as an AM process, computer-controlled axes in x- and y- direction for collector movement are essential. Furthermore, the printing head including heater, syringe and HV electrode with spinneret, is mounted on a manually driven z-axis for adjustment of the spinning length.
Figure 1:

Schematic of a MEW device.

In order to use MEW as an AM process, computer-controlled axes in x- and y- direction for collector movement are essential. Furthermore, the printing head including heater, syringe and HV electrode with spinneret, is mounted on a manually driven z-axis for adjustment of the spinning length.

Electrospinning configuration

Melt electrospinning was performed at 21±2 °C with a relative humidity of 40±10%. The polymer melt temperature for standard conditions was 73±1 °C unless specifically stated otherwise and the polymer was used between 1 and 5 days. Fibre pulsing regimes were investigated at different flow rate conditions using a 22 G spinneret at 0.2, 0.5, 1.0, and 2.0 bar. These experiments were performed with 6.0 kV acceleration voltage. Acceleration voltage refers here to a difference between negative voltage on the collector, which was always set to –1.5 kV and a variable positive spinneret voltage. In addition, the effect of the acceleration voltage at 4.50, 5.25 and 6.00 kV was investigated at 1.0 bar pressure. The distance between the spinneret tip and the collector surface (spinning length) was maintained at 4 mm for all experiments, if not specifically stated otherwise.

Digital signals in the form of G-codes were generated to enable the MEW device to print different lines of fibres in alternating 0° and 90° directions to form grids, as well as fibre line arrays for diameter quantification. Printing was done on a collector consisting of glass slides (26×50 mm) for fibre diameter measurements and stainless steel plates for scaffold processing (50×70 mm).

Imaging

Electrified PCL jet images were obtained using an AxioCam 105 colour digital camera (Carl Zeiss Microscopy, Göttingen, Germany), mounted on a 75–300 mm lens (Kenko Tokina Co., Ltd., Tokyo, Japan) outside the polycarbonate housing aimed at the spinneret. In order to analyse printed fibre lines and grid structures a stereomicroscope (Discovery V20, Carl Zeiss Microscopy GmbH, Göttingen, Germany) was used. SEM of the printed grids and scaffolds was performed using a Crossbeam 340 SEM (Carl Zeiss Microscopy, Göttingen, Germany) equipped with a Zeiss Gemini column. Quantitative SEM imaging of grids and line arrays was performed with samples as printed, with no coating applied to the surfaces, while qualitative images of scaffolds were sputter coated with platinum. SEM images were taken of specific regions of the printed grids, line arrays and edges to investigate the morphology of the printed fibres. SEM analysis was performed to determine diameters of the fibres by systematically measuring fibre diameters for the printed line arrays. Furthermore, for each varied parameter three repeats of the line array samples were analysed with 20 alternate fibres measured on each sample. The edges of both the grids and line arrays provided information related to the turn points.

Results and discussion

In this study, we assume a conservation of polymer masses with m1+m2+m3=constant, where m1 is the polymer melt in the spinneret (including reservoir), m2 the mass of the jet and m3 the mass of the deposited fibres. Figure 2 illustrates the corresponding mass flows of the MEW process schematically. During every single printing experiment, the polymer melt mass flow dm1/dt to the spinneret was kept constant (d2m1/dt=0). To investigate the oscillation of the resulting fibre mass flow dm3/dt we characterised deposited fibre diameters manufactured at constant collector speeds. This procedure enables an evaluation of a time-dependent fluctuation of the jet with dm2/dt. During the experiments no jet breaks were observed, even under pulsing or long beading conditions.

Schematic of mass flows involved in MEW processing. The molten polymer is fed through a spinneret via gas pressure with dm1/dt. The HV electrode (including spinneret and melt) becomes charged, as result a Taylor Cone forms and a jet is accelerated to the opposing collector with dm2/dt and finally deposited as a fibre with dm3/dt.
Figure 2:

Schematic of mass flows involved in MEW processing.

The molten polymer is fed through a spinneret via gas pressure with dm1/dt. The HV electrode (including spinneret and melt) becomes charged, as result a Taylor Cone forms and a jet is accelerated to the opposing collector with dm2/dt and finally deposited as a fibre with dm3/dt.

Effect of collector speed on fibre morphology

In MEW, straight fibres are printed when the computer-controlled collector movement speed vcol is greater than the ultimate jet speed, termed the critical translation speed (CTS) as described previously [12] (Figure 3A). When processing below this CTS value, non-linear fluid patterns due to jet buckling can be observed (Figure 3B–D). These types of patterns are already described for falling viscous liquids (syrup, silicone oil) under non-electrohydrodynamic conditions [39], [40], [41]. Such patterns have been described as: sinusoidal meanders (Figure 3B), having “sidekicks” or translated coiling (Figure 3C), and “figure of eight” loops (Figure 3D). The difference is that here, the molten PCL fibre solidifies due to cooling [42] while the syrup and silicone fluids coalesce and lose their shape. The uniform patterns shown for PCL in Figure 3 demonstrate that the molten electrified jet is stable with time, since any slight change in the diameter would affect the deposition behaviour.

Fabrication of different patterns by changes in collector speed at and below CTS. The fibres were deposited in a series of 10 lines on the collector with a 30 G spinneret at 2.5 kV acceleration voltage, 90 °C heating temperature 1.0 mm collector distance and 1.0 bar pressure. Fibre collection was performed at (A) 100%–110% of CTS, (B) 75%–80% of CTS, (C) 30%–35% of CTS and (D) 10%–15% of CTS.
Figure 3:

Fabrication of different patterns by changes in collector speed at and below CTS.

The fibres were deposited in a series of 10 lines on the collector with a 30 G spinneret at 2.5 kV acceleration voltage, 90 °C heating temperature 1.0 mm collector distance and 1.0 bar pressure. Fibre collection was performed at (A) 100%–110% of CTS, (B) 75%–80% of CTS, (C) 30%–35% of CTS and (D) 10%–15% of CTS.

At collector speeds above the CTS, the fibre diameter can be reduced by a high ratio of vcol/CTS (Figure 4A–C) due to mechanical stretching. Another effect of increasing collector speed is a growing lag of the jet. The first contact point of the fibre on the collector lags behind the location of the spinneret, as the collector moves and drags the electrified molten jet. This is especially visible on the deposited fibre at turning points (Figure 4A). The magnitude of this lag defines the discrepancy between the programmed collector movement and the deposited fibre path. Figure 4D–G shows the electrified molten jet being drawn along the surface of the collector at different speeds above the CTS.

Influence of the collector speed equal to CTS or greater on the fibre deposition and shape. (A) When vcol/CTS is increased the variable lag is increased resulting in a visible offset on the turning points at the grid edges. (B and C) In addition, the fibre diameter is further thinned by mechanical drawing. Experiments were performed with a 23 G spinneret and 5.7 kV, 4.4 mm, 3.0 bar at 82 °C. (D–G) Photographical imaging of the PCL jet during MEW: An increased collector speed led to an increased variable lag as indicated by the red arrows D: 500 mm/min (1.8×CTS), (E) 1000 mm/min (3.6× CTS), (F) 2000 mm/min (7.1×CTS) (G) 8000 mm/min (28.6×CTS). Experiments were performed with a 22 G spinneret and 6.5 kV, 6 mm, 1.0 bar at 73 °C. All scale bars are 1 mm; CTS=280±10 mm/min.
Figure 4:

Influence of the collector speed equal to CTS or greater on the fibre deposition and shape.

(A) When vcol/CTS is increased the variable lag is increased resulting in a visible offset on the turning points at the grid edges. (B and C) In addition, the fibre diameter is further thinned by mechanical drawing. Experiments were performed with a 23 G spinneret and 5.7 kV, 4.4 mm, 3.0 bar at 82 °C. (D–G) Photographical imaging of the PCL jet during MEW: An increased collector speed led to an increased variable lag as indicated by the red arrows D: 500 mm/min (1.8×CTS), (E) 1000 mm/min (3.6× CTS), (F) 2000 mm/min (7.1×CTS) (G) 8000 mm/min (28.6×CTS). Experiments were performed with a 22 G spinneret and 6.5 kV, 6 mm, 1.0 bar at 73 °C. All scale bars are 1 mm; CTS=280±10 mm/min.

Temporary pulsing

Different types of pulsing can be observed and are related to the duration of printing, Figure 5 gives a general explanation of the pulsing role in the jet geometry. When the electrified jet diameter varies from a thick to a thin profile (and vice versa), there is a visible change of the variable lag. Changes in the variable lag have a particular impact on fibre placement when turning during writing and therefore affects the scaffold quality after printing multiple layers.

Influence of pulsing on the variable lag. (A) Schematic of pulsing of the jet due to non-balanced process parameters including the variable lag as a function of jet thickness. (B) When depositing fibres above the CTS, pulsing can be readily observed as wave-like edges on the printed grids. Because temporary pulsing can be stabilised over time, the edge structure can be observed as dampened oscillation in this case. A regular G-code pattern was used to print the actual PCL sample shown here in (B). The photograph was taken on the stereomicroscope with dark-field background, and afterwards white/black inverted.
Figure 5:

Influence of pulsing on the variable lag.

(A) Schematic of pulsing of the jet due to non-balanced process parameters including the variable lag as a function of jet thickness. (B) When depositing fibres above the CTS, pulsing can be readily observed as wave-like edges on the printed grids. Because temporary pulsing can be stabilised over time, the edge structure can be observed as dampened oscillation in this case. A regular G-code pattern was used to print the actual PCL sample shown here in (B). The photograph was taken on the stereomicroscope with dark-field background, and afterwards white/black inverted.

Accordingly, we observed a temporary jet oscillation initiated by changes of instrumental parameters (e.g. feeding pressure) which seemed to shift the dynamic process and thus force balance to another equilibrium level. This shift in balance of parameters can cause a “temporary pulsing” instability, which leads to an initial disequilibrium of mass flows that equilibrates to a homogenous deposition profile (Figure 5B). We observed that such temporary jet oscillations can last several minutes before stabilising of the jet, as long as the combination of instrumental parameters is adjusted for stable printing.

The variable lag of the jet is of particular interest to the controlled placement of fibres during MEW. A balance of the instrumental parameters with respect to the influence of the jet speed on jet thickness is crucial in obtaining controlled scaffold manufacture. As shown in Figure 5A, a larger mass flow of the polymer melt through the spinneret (dm1/dt) not only led to thicker fibres, but also to a lower CTS [12]. As a result, the pulsing phenomenon can be observed as a wave-like structural feature at the turn edge (Figure 5B), given that MEW operates above the CTS throughout. This is a simple visual tool to evaluate if optimization of instrumental parameters is still required. The effect of changes in the jet profile due to pulsing can be compared to the programmed path of a regular deposition pattern. In case where the new equilibrium state cannot be achieved, for instance due to insufficient electrical field strength, wave-like edge is preserved throughout the constructs (Figure S1). This denotes such processing phenomenon as continuous pulsing.

Continuous pulsing

The instrumental parameters (22 G; 5.25 kV, 4 mm collector distance and 1.0 bar) for the demonstration of continuous pulsing were selected as they provided excessive flow rate conditions (dm1/dt) that resulted in continuous oscillation of fibre diameters in the following experiments. We previously described this effect as “fibre pulsing” for the MEW processing of PEtOx [15]. To demonstrate that fibre pulsing is also applicable to PCL, we investigated this phenomenon further. As illustrated in Figure 2, the polymer melt mass flow dm1/dt through the spinneret is a key component of fibre pulsing and is largely governed by the feeding pressure, melt viscosity and spinneret diameter/length, according to the Hagen-Poiseuille equation [43], [44]. The described correlation in this formula is valid for viscous, Newtonian fluids and laminar flow conditions and therefore does not fully apply for polymer melts, but nonetheless the equation can be helpful to assess the influence of instrumental parameters during MEW. When the spinneret size is constant during processing, the feeding pressure can be used to influence dm1/dt. Moreover, the feeding pressure seems to be the most suitable parameter for changing dm1/dt due to its broad adjustable range (here we used one order of magnitude: 0.2–2.0 bar) without additional consequences or limitations for the printing process beside pulsing phenomena.

Increased pressure can be compensated by greater magnitude of forces, stretching the jet, induced by an electrical field. Nevertheless, an excessive voltage leads to arcing and a breakdown of the electrical field. In our experience this resulted in a narrow usable voltage range at a fixed collector distance, while the acceleration voltage needs to be at a certain minimum value in order to form a Taylor Cone [13], Figure 7 shows that the increase in acceleration voltage from 5.25 kV to 6.00 kV resulted in more homogeneously printed fibres with 15.56±1.16 μm to 15.50±0.46 μm, respectively. According to the printing conditions with d2m1/dt=0, a stronger electrical field tended to stabilise the jet mass flow dm2/dt at the studied conditions. Hence, parameters which define forces affecting dm1/dt and dm2/dt must be balanced in order to avoid continuous pulsing with d2m2/dt≠0. Under pulsing conditions, the resulting force value which leads to jet stretching (as a function of Coulomb interaction, mechanical stretching, surface tension and counteracting restoring forces of the viscoelastic polymer etc.) cannot be constant during pulsing.

Long bead defects

For the samples produced at 2.0 bar (Figure 6A, B1–B2) and 4.5 kV (Figure 7A, D1–D2) the fibre diameter oscillation was even more evident than continuous pulsing. This phenomenon was termed here “long beading” which describes the shape of the printed instability. While both pulsing and long beading indicate an oscillation of the jet profile and can be observed at the edges of a printed construct, long beads can be characterised as outliers in the sorted function of fibre diameters at sufficient number of measurement points (n=60). In the case of 1.0 bar (Figure 6C1–C2), 0.5 bar (Figure 6D1–D2), and 0.2 bar (Figure 6E1–E2), no remarkable fibre oscillation of dm3/dt occurred, and homogeneous fibre diameters (15.40±0.60, 11.04±0.41, and 7.36±0.35 μm respectively, Table 1) (average±SD) could be observed (Figure 8). This is expressed by small deviation with median and averages of measured fibre diameters being very similar. In contrast, the 2.0 bar samples contained long beads and had comparatively large deviations recorded for fibre diameter measurements (19.4±8.4 μm) with the measured diameters ranging from 8.0 μm up to 46.4 μm (Figure 8B). A visualisation of a representative long bead is provided in the supplementary information (Figure S2).

Printed structures at different feeding pressures. (A) Images of the collector plate after MEW. The collected material was deposited in three zones – stabilisation area, grids and fibres for diameter measurements. While the feeding pressure was set to (B) 2.0 bar, (C) 1.0 bar, (D) 0.5 bar, and (E) 0.2 bar, other parameters were kept constant. The fibre diameters increased visibly with increasing pressure. It can be observed in the images that long beading and pulsing occurred as a result of excessive feeding pressure (2.0 bar). This resulted also in irregular deposited edges of the lines for diameter measurement. The black arrow in B2 points to a thicker fibre segment with a higher variable lag due to long beading. White arrows in B1–E2 show the printing direction. SEM: scale bars=100 μm.
Figure 6:

Printed structures at different feeding pressures.

(A) Images of the collector plate after MEW. The collected material was deposited in three zones – stabilisation area, grids and fibres for diameter measurements. While the feeding pressure was set to (B) 2.0 bar, (C) 1.0 bar, (D) 0.5 bar, and (E) 0.2 bar, other parameters were kept constant. The fibre diameters increased visibly with increasing pressure. It can be observed in the images that long beading and pulsing occurred as a result of excessive feeding pressure (2.0 bar). This resulted also in irregular deposited edges of the lines for diameter measurement. The black arrow in B2 points to a thicker fibre segment with a higher variable lag due to long beading. White arrows in B1–E2 show the printing direction. SEM: scale bars=100 μm.

Printed structures at different acceleration voltages. (A) Images of the collector plate after MEW. The collected material was deposited in three zones – stabilisation zone, grids and fibres for diameter measurements. While the acceleration voltage was set to (B) 6.0 kV, (C) 5.25 kV, and (D) 4.5 kV, other parameters were kept constant. As it can be observed in the images, long beading (4.50 kV) and pulsing (5.25 kV) occurred at lower voltages. White arrows in B1–D2 show the printing direction. SEM: scale bar=100 μm.
Figure 7:

Printed structures at different acceleration voltages.

(A) Images of the collector plate after MEW. The collected material was deposited in three zones – stabilisation zone, grids and fibres for diameter measurements. While the acceleration voltage was set to (B) 6.0 kV, (C) 5.25 kV, and (D) 4.5 kV, other parameters were kept constant. As it can be observed in the images, long beading (4.50 kV) and pulsing (5.25 kV) occurred at lower voltages. White arrows in B1–D2 show the printing direction. SEM: scale bar=100 μm.

Table 1:

Fibre diameter distribution at different instrumental parameters.

Fibre diameters as a function of feeding pressures and acceleration voltages. (A and B) With increased feeding pressures, the fibre diameter increased. Long beading was observed at high pressures (2.0 bar) under the chosen processing conditions. (C and D) With decreasing acceleration voltage from 6.00 to 4.50 kV fibre pulsing and finally long beading occurred.
Figure 8:

Fibre diameters as a function of feeding pressures and acceleration voltages.

(A and B) With increased feeding pressures, the fibre diameter increased. Long beading was observed at high pressures (2.0 bar) under the chosen processing conditions. (C and D) With decreasing acceleration voltage from 6.00 to 4.50 kV fibre pulsing and finally long beading occurred.

The fibre diameter for both experimental series (pressure and voltage) can be analysed with the graphical application of the average in combination with the median. This seems to be a simple tool to analyse the influence of long beading during MEW, since a homogenous jet diameter results in a narrow and symmetrical distribution function (Figure 8) with a ratio of median/average (m/a) close to 1.0. In contrast, long beading as a result of high pressures (2.0 bar) or low voltages (4.50 kV) was expressed by a difference of 14% (m/a=0.86) and 25% (m/a=0.75), respectively. Thus, the asymmetry of the distribution functions indicates the appearance and the intensity of long beading.

Printing instabilities

Studying the mass flows and visualisation of the process can provide a better understanding of process stability during MEW. A decrease in acceleration voltage at constant dm1/dt led to time-dependant oscillation of jet and fibre mass flows with d2m2/dt≠0 and thus d2m3/dt≠0. A similar effect can be observed when the feeding pressure is increased under constant electrical field strength. Then an increased dm1/dt led to an excessive material extrusion, initiating the time dependant fibre oscillation of d2m3/dt≠0 (continuous pulsing or long beading). The electrical field induces a force acting on the molten polymer jet leading to melt acceleration and therefore influences the stability of dm2/dt (electrified jet mass). If the acceleration voltage was not sufficient or the feeding pressure too high, then continuous pulsing or long beading occurred. We hypothesise that the capacity of the electrical field to accelerate the mass to the collector was exceeded by a high extrusion rate through the spinneret. After a certain time, forces acting on the accumulated polymer at the spinneret tip overcome the mechanical resistance of the material. Then, the melt is drawn to the collector in the form of a thicker jet, preserving the Taylor cone (pulsing) or as an elongated drop (long beading). Afterwards the jet gets thinner and CTS increases. This behaviour in terms of fibre diameter change over time is illustrated in Figure 9, based on the measurements presented in the supplementary information (Figure S3).

Schematic of the deposited fibre mass flow. (A) Fibre mass flow dm3/dt based on diameter measurements shown in Supplementary Figure S3. (B) Deposited fibres are illustrated. Despite the stable polymer flow through the spinneret dm1/dt, inhomogeneous fibres were printed due to insufficient acceleration voltages. Thereby an oscillation of dm2/dt led to pulsing and long beading.
Figure 9:

Schematic of the deposited fibre mass flow.

(A) Fibre mass flow dm3/dt based on diameter measurements shown in Supplementary Figure S3. (B) Deposited fibres are illustrated. Despite the stable polymer flow through the spinneret dm1/dt, inhomogeneous fibres were printed due to insufficient acceleration voltages. Thereby an oscillation of dm2/dt led to pulsing and long beading.

Comparison with instabilities during melt spinning

There are already different manufacturing instabilities described for common melt spinning including draw resonance and melt fracture which also lead to a fibre diameter oscillation. As in the case of melt spinning, molten polymers are pressed through spinnerets but in contrast to MEW, an uptake wheel draws the solidifying polymer fibres by mechanical stretching only. For melt spinning, the draw resonance is usually described as an instability occurring if a critical drawdown ratio (defined as fibre uptake speed divided by melt extrusion speed) is exceeded, ultimately leading to fibre breakage [45], [46], [47], [48]. In this context, draw resonance was also described as partially disappearing at a second stable region or at high drawdown ratios, also depending on the spinning length (distance between spinnerets and uptake wheels for melt spinning) [49], [50], [51], [52], [53]. Nevertheless, draw resonance cannot be readily compared to MEW due to the vastly different process conditions. MEW involves a different experimental set-up including a relatively small spinning length (collector distance) and the application of an accelerating voltage that both draws the jet and stabilises the fluid from jet break up. Most importantly, the very low mass flow rate required to establish a stable MEW process – typically 20–50 μL/h [15], [22] and as low as 5 μL/h [12] is significantly lower compared to industrialised techniques such as melt spinning. Thus, the comparison of these two unstable melt operating regimes is speculative without further investigation.

Stable printing conditions

In order to print regularly structured scaffolds with MEW, it is important to avoid continuous pulsing or long beading (Figure 10). Due to changes in the CTS during fibre diameter oscillation, the scaffold quality decreases, especially at the turning loops. The jet and fibre drag is influenced by the homogeneity of the jet diameter and can lead to inaccurate fibre deposition. Figure 10A–B, G–H shows defective scaffold structures as a result of changes in printing direction during long beading conditions. Obviously, this instability needs to be prevented to enable stable printing conditions. Indeed, the feeding pressure and acceleration voltage effect on the deposition quality was shown, however, the influence of polymer characteristics, processing temperature, collector distance, spinneret size and environmental conditions should be investigated in order to optimise MEW over a broader range of conditions.

SEM images of the scaffolds. They were made using 2×10 layers of the grid patterns. Perpendicular and tilted SEM images are shown for each group. (A–B) 2.0 bar; 6.00 kV. (C–D) 1.0 bar; 6.00 kV. (E–F) 1.0 bar; 5.25 kV. (G–H) 1.0 bar; 4.5 kV. (I–J) 0.5 bar; 6.00 kV. (K–L) 0.2 bar; 6.00 kV. Long beading was observed in A, B, G and H while continuous pulsing occurred in E and F.
Figure 10:

SEM images of the scaffolds.

They were made using 2×10 layers of the grid patterns. Perpendicular and tilted SEM images are shown for each group. (A–B) 2.0 bar; 6.00 kV. (C–D) 1.0 bar; 6.00 kV. (E–F) 1.0 bar; 5.25 kV. (G–H) 1.0 bar; 4.5 kV. (I–J) 0.5 bar; 6.00 kV. (K–L) 0.2 bar; 6.00 kV. Long beading was observed in A, B, G and H while continuous pulsing occurred in E and F.

Conclusions

Fibre pulsing during MEW is a common phenomenon that occurs due to a force imbalance leading to mass flow oscillations and therefore to inhomogeneous fibre diameters. Since MEW has multiple instrumental parameters that affect fibre collection, balancing the parameters responsible for the jet stability can be challenging. Consistent diameters of the fibres are important for the fabrication of precise scaffolds, since jet homogeneity influences fibre placement. We demonstrate that pulsing can be observed by utilising the difference in the variable lag that follows the oscillation of jet thickness at the edge of simple collector movement paths. Long beading is a result of an extreme oscillation of mass flows that can be analysed by sorting and evaluating a sufficient number of measured fibre diameters. In order to avoid these undesired process instabilities, it is suitable to limit the mass flow through the spinneret by decreasing the feeding pressure or by stabilising the mass flow of the polymer jet by increasing the electrical field strength. Our presented approach to the operation of a MEW device should allow the user to identify pulsing and diminish this phenomenon to improve scaffold quality.

Acknowledgments

We would like to thank the European Research Council (ERC) (consolidator grant Design2Heal, contract # 617989) for their financial support and the Australian Technology Network - German Academic Exchange Service (ATN-DAAD) for their support of JNH. The authors would like to thank Dr. Claus Moseke, Ms. Judith Friedlein, and Mr. Joachim Liebscher for the technical assistance. TJ designed and assembled the MEW device. Further, we gratefully acknowledge financial support by the German Research Foundation (DFG) State Major Instrumentation Programme, for funding the Zeiss Crossbeam CB 340 scanning electron microscope (INST 105022/58-1 FUGG).

References

  • 1.

    Hutmacher DW, Woodfield TB, Dalton PD. Scaffold design and fabrication. In: van Blitterswijk C, De Boer J, editors. Tissue Engineering, 2nd ed. Amsterdam: Elsevier, 2014. ISBN 9780124201453, pp. 311–46. Google Scholar

  • 2.

    Dalton PD, Woodfield T, Hutmacher DW. SnapShot: polymer scaff olds for tissue engineering. Biomaterials. 2009;30:701–2. Google Scholar

  • 3.

    Karp JM, Dalton PD, Shoichet MS. Scaffolds for tissue engineering. Mrs Bull. 2003;28:301–6. Google Scholar

  • 4.

    Karp JM, Rzeszutek K, Shoichet MS, Davies JE. Fabrication of precise cylindrical three-dimensional tissue engineering scaffolds for in vitro and in vivo bone engineering applications. J Craniofac Surg. 2003;14:317–23. Google Scholar

  • 5.

    Teng YD, Lavik EB, Qu XL, Park KI, Ourednik J, Zurakowski D, et al. Correction. P Natl Acad Sci USA. 2002;99:9606. Google Scholar

  • 6.

    Malda J, Visser J, Melchels FP, Jungst T, Hennink WE, Dhert WJ, et al. 25th Anniversary Article: engineering hydrogels for biofabrication. Adv Mater. 2013;25:5011–28. Google Scholar

  • 7.

    Melchels FP, Domingos MA, Klein TJ, Malda J, Bartolo PJ, Hutmacher DW. Additive manufacturing of tissues and organs. Prog Polym Sci. 2012;37:1079–104. Google Scholar

  • 8.

    Hutmacher DW, Schantz T, Zein I, Ng KW, Teoh SH, Tan KC. Mechanical properties and cell cultural response of polycaprolactone scaffolds designed and fabricated via fused deposition modeling. J Biomed Mater Res. 2001;55:203–16. Google Scholar

  • 9.

    Melchels FP, Bertoldi K, Gabbrielli R, Velders AH, Feijen J, Grijpma DW. Mathematically defined tissue engineering scaffold architectures prepared by stereolithography. Biomaterials. 2010;31:6909–16. Google Scholar

  • 10.

    Skoog SA, Goering PL, Narayan RJ. Stereolithography in tissue engineering. J Mater Sci-Mater M. 2014;25:845–56. Google Scholar

  • 11.

    Dalton PD, Vaquette C, Farrugia BL, Dargaville TR, Brown TD, Hutmacher DW. Electrospinning and additive manufacturing: converging technologies. Biomater Sci-UK. 2013;1:171–85. Google Scholar

  • 12.

    Brown TD, Dalton PD, Hutmacher DW. Direct writing by way of melt electrospinning. Adv Mater. 2011;23:5651–7. Google Scholar

  • 13.

    Taylor G. Electrically driven jets. Proc Roy Soc Lond A Mat. 1969;313:453–75. Google Scholar

  • 14.

    Zein I, Hutmacher DW, Tan KC, Teoh SH. Fused deposition modeling of novel scaffold architectures for tissue engineering applications. Biomaterials. 2002;23:1169–85. Google Scholar

  • 15.

    Hochleitner G, Hummer JF, Luxenhofer R, Groll J. High definition fibrous poly(2-ethyl-2-oxazoline) scaffolds through melt electrospinning writing. Polymer. 2014;55:5017–23. Google Scholar

  • 16.

    Brown TD, Slotosch A, Thibaudeau L, Taubenberger A, Loessner D, Vaquette C, et al. Design and fabrication of tubular scaffolds via direct writing in a melt electrospinning mode. Biointerphases. 2012;7:13. Google Scholar

  • 17.

    Hochleitner G, Jungst T, Brown TD, Hahn K, Moseke C, Jakob F, et al. Additive manufacturing of scaffolds with sub-micron filaments via melt electrospinning writing. Biofabrication. 2015;7:035002. Google Scholar

  • 18.

    Visser J, Melchels FP, Jeon JE, van Bussel EM, Kimpton LS, Byrne HM, et al. Reinforcement of hydrogels using three-dimensionally printed microfibres. Nat Commun. 2015;6:6933. Google Scholar

  • 19.

    Muerza-Cascante ML, Haylock D, Hutmacher DW, Dalton PD. Melt electrospinning and its technologization in tissue engineering. Tissue Eng. 2015;21:187–202. Google Scholar

  • 20.

    Thibaudeau L, Taubenberger A, Holzapfel BM, Quent VM, Fuehrmann T, Hesami P, et al. A tissue-engineered humanized xenograft model of human breast cancer metastasis to bone. Dis Mod Mech. 2014;7:299–309. Google Scholar

  • 21.

    Farrugia BL, Brown TD, Upton Z, Hutmacher DW, Dalton PD, Dargaville TR. Dermal fibroblast infiltration of poly(ε-caprolactone) scaffolds fabricated by melt electrospinning in a direct writing mode. Biofabrication. 2013;5:025001. Google Scholar

  • 22.

    Brown TD, Edin F, Detta N, Skelton AD, Hutmacher DW, Dalton PD. Melt electrospinning of poly(ε-caprolactone) scaffolds: phenomenological observations associated with collection and direct writing. Mater Sci Eng C Mater Biol Appl. 2014;45:698–708. Google Scholar

  • 23.

    Jungst T, Muerza-Cascante ML, Brown TD, Standfest M, Hutmacher DW, Groll J, et al. Melt electrospinning onto cylinders: effects of rotational velocity and collector diameter on morphology of tubular structures. Polym Int. 2015;64:1086–95. Google Scholar

  • 24.

    Kim SJ, Jang DH, Park WH, Min BM. Fabrication and characterization of 3-dimensional PLGA nanofiber/microfiber composite scaffolds. Polymer. 2010;51:1320–7. Google Scholar

  • 25.

    Zhou HJ, Green TB, Joo YL. The thermal effects on electrospinning of polylactic acid melts. Polymer. 2006;47:7497–505. Google Scholar

  • 26.

    Dalton PD, Klinkhammer K, Salber J, Klee D, Moller M. Direct in vitro electrospinning with polymer melts. Biomacromolecules. 2006;7:686–90. Google Scholar

  • 27.

    Dalton PD, Lleixa Calvet J, Mourran A, Klee D, Moller M. Melt electrospinning of poly-(ethylene glycol-block-ε-caprolactone). Biotechnol J. 2006;1:998–1006. Google Scholar

  • 28.

    Dalton PD, Grafahrend D, Klinkhammer K, Klee D, Moller M. Electrospinning of polymer melts: phenomenological observations. Polymer. 2007;48:6823–33. Google Scholar

  • 29.

    Dalton PD, Joergensen NT, Groll J, Moeller M. Patterned melt electrospun substrates for tissue engineering. Biomed Mater. 2008;3:034109. Google Scholar

  • 30.

    Detta N, Brown TD, Edin FK, Albrecht K, Chiellini F, Chiellini E, et al. Melt electrospinning of polycaprolactone and its blends with poly(ethylene glycol). Polym Int. 2010;59:1558–62. Google Scholar

  • 31.

    Chen F, Hochleitner G, Woodfield T, Groll J, Dalton P, Amsden BG. Additive manufacturing of a photo-cross-linkable polymer via direct melt electrospinning writing for producing high strength structures. Biomacromolecules. 2015;17:208–14. Google Scholar

  • 32.

    Lyons J, Li C, Ko F. Melt-electrospinning part I: processing parameters and geometric properties. Polymer. 2004;45:7597–603. Google Scholar

  • 33.

    Wang XF, Huang ZM. Melt-electrospinning of PMMA. Chinese J Polym Sci. 2010;28:45–53. Google Scholar

  • 34.

    Fridrikh SV, Yu JH, Brenner MP, Rutledge GC. Controlling the fiber diameter during electrospinning. Phys Rev Lett. 2003;90:144502. Google Scholar

  • 35.

    Reneker DH, Yarin AL. Electrospinning jets and polymer nanofibers. Polymer. 2008;49:2387–425. Google Scholar

  • 36.

    Wei C, Dong JY. Direct fabrication of high-resolution three-dimensional polymeric scaffolds using electrohydrodynamic hot jet plotting. J Micromech Microeng. 2013;23:025017. Google Scholar

  • 37.

    Woodruff MA, Hutmacher DW. The return of a forgotten polymer – polycaprolactone in the 21st century. Prog Polym Sci. 2010;35:1217–56. Google Scholar

  • 38.

    Probst FA, Hutmacher DW, Muller DF, Machens HG, Schantz JT. Rekonstruktion der Kalvaria durch ein präfabriziertes bioaktives Implantat. Handchir Mikrochir P. 2010;42:369–73. Google Scholar

  • 39.

    Chiu-Webster S, Lister JR. The fall of a viscous thread onto a moving surface: a ‘fluid-mechanical sewing machine’. J Fluid Mech. 2006;569:89–111. Google Scholar

  • 40.

    Welch RL, Szeto B, Morris SW. Frequency structure of the nonlinear instability of a dragged viscous thread. Phys Rev E. 2012;85:066209. Google Scholar

  • 41.

    Brun PT, Audoly B, Ribe NM, Eaves TS, Lister JR. Liquid ropes: a geometrical model for thin viscous jet instabilities. Phys Rev Lett. 2015;114:174501. Google Scholar

  • 42.

    Zhmayev E, Cho D, Joo YL. Electrohydrodynamic quenching in polymer melt electrospinning. Phys Fluids. 2011;23:073102. Google Scholar

  • 43.

    Pfitzner J. Poiseuille and his law. Anaesthesia. 1976;31:273–5. Google Scholar

  • 44.

    Sutera SP, Skalak R. The history of poiseuille’s law. Annu Rev Fluid Mech. 1993;25:1–19. Google Scholar

  • 45.

    Bergonzoni A, DiCresce A. The phenomenon of draw resonance in polymeric melts. Part I – qualitative view. Polym Eng Sci. 1966;6:45–9. Google Scholar

  • 46.

    Jung HW, Hyun JC. Instabilities in extensional deformation polymer processing. Rheology Reviews. 2006;2006:131. Google Scholar

  • 47.

    Kase S, Matsuo T. Studies on melt spinning. II. Steady-state and transient solutions of fundamental equations compared with experimental results. J Appl Polym Sci. 1967;11:251–87. Google Scholar

  • 48.

    Kase S. Studies on melt spinning. IV. On the stability of melt spinning. J Appl Polym Sci. 1974;18:3279–304. Google Scholar

  • 49.

    Lamb P. Analysis of fabrication processes. SCI Monograph. 1967;26:296. Google Scholar

  • 50.

    Ishihara H, Kase S. Studies on melt spinning. VI. Simulation of draw resonance using Newtonian and power law viscosities. J Appl Polym Sci. 1976;20:169–91. Google Scholar

  • 51.

    Fisher RJ, Denn MM. A theory of isothermal melt spinning and draw resonance. Aiche J. 1976;22:236–46. Google Scholar

  • 52.

    Lee WK. A correlation of the onset of draw resonance spinline instability. Chem Eng Commun. 1987;53:117–30. Google Scholar

  • 53.

    Petrie CJ, Denn MM. Instabilities in polymer processing. Aiche J. 1976;22:209–36. Google Scholar

Supplemental Material:

The online version of this article (DOI: 10.1515/bnm-2015-0022) offers supplementary material, available to authorized users.

About the article

Received: 2015-12-17

Accepted: 2016-04-08

Published Online: 2016-05-19

Published in Print: 2016-09-01


Author’s statement

Conflict of interest: Authors state no conflict of interest.

Materials and methods

Informed consent: Informed consent has been obtained from all individuals included in this study.

Ethical approval: The research related to human use has been complied with all the relevant national regulations, institutional policies and in accordance the tenets of the Helsinki Declaration, and has been approved by the authors’ institutional review board or equivalent committee.


Citation Information: BioNanoMaterials, ISSN (Online) 2193-066X, ISSN (Print) 2193-0651, DOI: https://doi.org/10.1515/bnm-2015-0022.

Export Citation

©2016, Paul D. Dalton et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. BY-NC-ND 3.0

Supplementary Article Materials

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

[1]
Sarah Bertlein, Daichi Hikimoto, Gernot Hochleitner, Julia Hümmer, Tomasz Jungst, Michiya Matsusaki, Mitsuru Akashi, and Jürgen Groll
Small, 2017, Page 1701521
[2]
Andrew C. Daly, Fiona E. Freeman, Tomas Gonzalez-Fernandez, Susan E. Critchley, Jessica Nulty, and Daniel J. Kelly
Advanced Healthcare Materials, 2017, Page 1700298
[3]
Paul D. Dalton
Current Opinion in Biomedical Engineering, 2017, Volume 2, Page 49
[4]
Shixuan Chen, Bing Liu, Mark A Carlson, Adrian F Gombart, Debra A Reilly, and Jingwei Xie
Nanomedicine, 2017, Volume 12, Number 11, Page 1335
[5]
Jodie N. Haigh, Tim R. Dargaville, and Paul D. Dalton
Materials Science and Engineering: C, 2017, Volume 77, Page 883
[6]
Xiao-Xiao He, Jie Zheng, Gui-Feng Yu, Ming-Hao You, Miao Yu, Xin Ning, and Yun-Ze Long
The Journal of Physical Chemistry C, 2017, Volume 121, Number 16, Page 8663
[7]
Almoatazbellah Youssef, Scott J Hollister, and Paul D Dalton
Biofabrication, 2017, Volume 9, Number 1, Page 012002

Comments (0)

Please log in or register to comment.
Log in