450 delegates attended a programme of 73 papers in 17 sessions, plus student papers, table-top displays and a new technology showcase
Renaissance Hotel - Baltimore. Venue for INTC 2003
Hurricane Isabel did her best to dampen spirits and many delegates were forced into early or late departures by flight cancellations. However about 100 remained through the last morning, and those who failed to get out at all that day had the unusual experience of watching Chesapeake Bay trying to check into the conference hotel on Friday morning.
Nevertheless, it was one of the best technical conferences of recent years, the highlight, for your correspondent at least, being the informative and coherent sessions on nanofibers.
Thomas Kehl, the New MD of Freudenberg implored the nonwovens industry to speed up new product development to allow faster growth. He was critical of the whole development process:
However in a word of encouragement for the smaller companies he observed that external sources of ideas tended to deliver better returns.
There were several traps for an unwary management board trying to grow a new business:
Ted Wirtz, the retiring President of Inda opened the discussion with a presentation of 2002 industry statistics:
In the 2000 to 2002 period 302,000 tonnes of new capacity have been added in NA:
Polymer to fabric processes now accounted for 37% of world capacity and would continue to grow at 6-8% per annum. New composite structures would drive this growth (“SMXMS” - the X being unspecified but pulp is likely to be important.)
Fiber to fabric processes accounted for the remainder and would grow at 3-5% per annum. This growth would be driven by spunlace wipes and air-laid absorbent cores.
Other key trend drivers would be:
In future the flow of new products would be the reverse of the traditional, i.e. the consumer product makers would develop the ideas and specify the nonwovens and raw materials to be used.
In conclusion, Mr Wirtz observed that too many nonwoven producers and raw material suppliers were too ready to accept commodity supplier status and cut back new product development. R&D really did pay off in this industry.
Lee Sullivan of Freudenberg Nonwovens focussed on the importance of the automotive market to the US economy in general - “35% of GDP is auto-related”, and to a polyester spunbond producer in particular. He stressed the highly technical nature of PET spunbond production c.f. PP spunbond and related Freudenberg’s continued success in this sector to the difficulty competitors had in meeting the ever-tighter requirements of the auto and carpet markets. For the future he saw a resurgence of use of carpet tiles, this time shaped to prevent the fact that they were tiles being so obvious.
Chuck Moestra of Rohm and Haas provided an outlook from the viewpoint of a binder and additive supplier. Opportunities for growth would come from
New performance requirements:
And new market opportunities in:
- But not in apparel or home furnishing.
The panel provided answers to some audience questions:
Chris Fandrey of Powerscope Inc introduced the Ensemble Diffraction method for scanning nonwoven webs to obtain real-time information on fiber diameters. The technique uses a laser beam which, after scattering on passage through the web, is analysed by a “ring detector” which measures the scattered beam’s intensity at various distances from its original axis.
Fraunhofer scattering equations are applied, suitably corrected to compensate for:
Results can be obtained every second, and the beam can be up to 12mm wide. It can be traversed across the web to give a full width fiber diameter profile.
Correlation with SEM measurements of fiber size is excellent, the scattering technique giving slightly higher diameter values due to fiber bundling. Used on-line in a melt-blowing process it picks up the diameter changes with distance from the nozzle and can detect shot levels as an anomalous scatter at small angles.
The technique is currently limited to webs with >35% light transmission, and this means webs lighter than about 4 gsm. Practical fabrics have to be “thinned out” for off-line measurement, but melt-blown fibers can be measured on-line in the nozzle-to-drum region.
In response to questions:
Dr Steve Russell of Leeds University (UK) described small-scale work on a hydroentanglement analogue of the Laroche - Napco “3D needlefelt” technology. Water jets had been used to bond two prebonded webs together either side of spacers, which, if tubular, could simultaneously be used to fill the spaces with powders. He thought the powder could be a superabsorbent if it were possible to dry the web in the space between the last HE zone and the end of the spacers.
Photo’s of single and two-side entangled products were shown, although any continuous operation on existing HE machines would not allow two sided entanglement without collapsing the tubes. One sample had a PU coat on the inside of the tube, presumably made by using an annular spacer.
Alfred Watzl of Fleissner GmbH & Co offered spunlace lines 7 m wide operating at 600 bar water pressure and running at 250m/min on carded products (200 if aperturing required) or 400m/min on spunbond. Composite manufacturing capabilities were based on Fleissner’s exclusive licence to use Georgia-Pacific’s 3 layer hydroentanglement technology, the key example being the spunbond/pulp/spunbond composite with 75% pulp and 10 gsm spunbonds produced at 400m/min.
The paper also mentioned the unique Nanoval “melt-blown” technology that uses a Laval nozzle and ambient air to draw and explode the forming filaments into a bundle of continuous microfibers of 0.03 to 0.35 dtex. The technique gives much higher throughput than conventional melt-blown (~100gms/min/hole) and produces stronger more oriented nonwovens.
Asked what the relationship between Fleissner and Nanoval was, Mr Watzl said there was no formal relationship but the technique was interesting and they were collaborating on its development. The process was not yet commercial.
(NB Since the conference, Neumag has acquired the Nanoval technology)
Dr Benham Pourdeyhimi of NCRC is running a series of experiments to investigate structure, process and property relationships in hydroentanglement. First results indicate that increasing hydroentanglement energy from 0.3 to 2.2 Kwh/kg does not affect MD/CD orientation in the web. However, in production processes, high pressures tend to stick the web to the forming belt, and the force needed to remove it would tend to increase the MD orientation. MD tensile strength and bending rigidity increases with bonding energy up to about 1.5 Kwh/kg, and then declines. Fabric elongation increases linearly with energy for nylon webs and is unaffected for polyester webs.
Fabric thickness decreases to a minimum around 1.5 kwh/kg and then increases, appearing to be roughly the inverse of the tensile trend.
Asked if the way the energy is applied affects the results, Dr Pourdeyhimi said it would. However, in practice it would not be possible to use either hundreds of nozzles at low pressure or one nozzle at very high pressure to investigate the effects.
Dr Krishna Gupta of Porous Materials Inc showed how their Water Intrusion Porosimeter (WIP), designed for use on hydrophobic structures, could be used in conjunction with the standard Mercury Intrusion Porosimeter (MIP) to give extra detail of pore structure in fabrics made of mixtures of hydrophilic and hydrophobic fibers. In the WIP test, the water spontaneously fills hydrophilic pores but only fills hydrophobic pores under pressure, while in the MIP test all pores require pressure to fill them. (Hg is universally non-wetting). So, in the WIP test the uptake/pressure curve is a measure of hydrophobic pores only and the MIP test uptake/pressure curve relates to all the pores. The difference relates to the hydrophilic pore structure. For an unspecified “fibrous paper”, 82% of the pore volume was found to be due to hydrophilic pores with an average size of 60 microns, and 18% of the pore volume was due to hydrophobic pores with an average size of 9 microns.
Asked about the pore size ranges detectable, Dr Gupta said 20 microns was currently the largest detectable by WIP, but a new technique was being developed to measure pores up to 1mm in diameter. MIP worked from 30 angstroms up to 400 microns but needed very high pressures. The advantage of WIP was its relative simplicity on hydrophobic materials.
Dr Hooman Tafreshi of NCRC has compared observations of water jets using high speed cameras with theoretical expectations based on fluid dynamics. When the cone-shaped nozzles have the high pressure water entering at the base (“cone-up”) the calculated Reynolds number (10,600 to 36,000) predicts that the jet should atomize over a wide pressure range, but in practice, this only occurs at pressures above 2000psi. If the nozzle is inverted with respect to the water flow - “cone down” – atomisation does not occur even at pressures above 2000psi. This has been shown to be due to the water riding on a cushion of air throughout the cone-down nozzle.
This cannot happen in the cone-up configuration for any practical nozzles, but does occur if the cone angle is greater than 60 degrees.
At pressures above 2700psi the water jet appears unstable, the atomisation point varying with time under apparently constant process conditions. Modelling suggests that this is caused by cavitation, the onset of which allows air into the nozzle and creates a “hydraulic flip”.
Asli Begenir now with Sara Lee, collaborated with Pourdeyhimi and Tafreshi at NCRC to produce this paper. They studied the effect of water pressure and nozzle design on the discharge coefficient, the water jet profile and its breakup length. One cone-shaped nozzle with an L/D of 7.65 was used both cone-up and cone-down relative to the high pressure side of the injector, and compared with a second cylindrical nozzle with an L/D of 1. Water pressure was varied from 100 to 5000 psi.
The conclusions were:
Marshall Oathout of Dupont Inc. presented new procedures for investigating the factors involved in wiping a wet surface clean whether liquid is present to assist cleaning or as the result of a spill. The factors were:
The dynamic wiping efficiency test was developed using 9 commercial clean room wipes. 5 were hydroentangled nonwovens containing cellulose, one was woven cotton, two were knitted polyester fabrics and one was a PP spunbond. Each fabric had its maximum absorbent capacity checked and dry samples were then used to wipe up a range of spill volumes (up to 130% of this capacity) on a stainless steel surface. The wipe was attached to a sled pulled at 25 cm/sec over the surface. Unsurprisingly, the hydrophilic wipes outperformed the hydrophobic ones, with the 100% lyocell HE wipe appearing best when used both dry and pre-wetted to close to its saturation point. In the pre-wetted test, woodpulp/PET HE fabrics performed almost as well as the lyocell.
The wet particle removal test was in essence the above dynamic wiping test, the spill being contaminated with 10million 1.6 micron latex particles. After the wiping test, the surface was rinsed and the resulting liquid passed through a particle counter to establish how many particles were left behind. In this test, the 100% lyocell HE fabric did best both dry and prewetted. Interestingly every wipe removed more particles than expected from the % of spill left behind, so the particles appear to preferentially stay with the wipe. This effect was most apparent for lyocell and least apparent for the knitted fabrics.
The “particles contributed by the wipe” test was the wet particle removal test using an uncontaminated spill. Particles in any spill left on the surface would therefore have come from the wipe. 1-3 micron particles were counted. Here the knitted Polyester and the 100% polyester HE appeared best with the 100% lyocell HE being good up to about 90% of its capacity. Particle release from all fabrics was much lower than expected from earlier static particle release testing.
Comparisons of wiping stainless steel, polyethylene and glass surfaces showed that glass was much more retentive to particles. Stainless and PE were similar.
Testing with 0.6 micron particles gave results very similar to the 1.6 micron particles.
Wiping iso-propyl alcohol spillages proved difficult on this rig because the spill spread out beyond the sled width for all but the smallest spill.
The bottom line: It’s necessary to wipe dry to wipe clean.
Nick Simpson of Tencel Ltd reviewed 10 years of lyocell development and summarised the attributes of the fiber across the whole range of nonwoven products.
Wipes had proved to be the most successful nonwoven end-use. Hydroentanglement and needling were the most used conversion technologies. Nonwovens using lyocell were better than viscose nonwovens for strength, especially wet strength, for aperturing clarity, softness (at comparable strength), basis weight range in spunlacing, and low-linting. In air-laying the low cohesion of the fiber gave good results, and in needlepunching the ability to use a 1.7 dtex fiber led to softer products.
In response to questions, Mr Simpson said available Tencel fiber types were restricted to between 1 and 3 denier, but any length from 0.5mm upwards could be supplied. Prices could not be discussed, but it was comparable with viscose and hence more expensive than polyester staple. Wicking and water retention properties were similar to viscose. The dissolution process did not degrade wood-pulp so fiber and pulp had similar degrees of polymerisation.
Ivo Edward Ruzek of Industrial Consulting estimated that worldwide use of polyester in nonwovens was now around 500,000 tonnes and felt that future growth needed a fiber specially tailored to the needs of the nonwoven producers. The traditional cotton-type fiber as sold to the nonwoven industry was no longer good enough:
So Mr Ruzek proposed a partially oriented fiber with a strength of about 4.2 gpd and 65% extension made from PET with ethylene or propylene glycol blocks to get better moisture management or with PLA to lower the melting point. Furthermore, if it could be made with a hollow core, weight and cost savings could follow while the hollow fiber would give improved resilience.
Diana Ortiz of the University of Oklahoma gave Robert Shambaugh’s paper on producing PP fibers with 50% hollowness, this being defined as (ID/OD) 2 where ID is the diameter of the hole, and OD is the outside diameter of the fiber. In a low-speed, short-spin (0.5 m threadline) spunbond process this could be achieved by injecting nitrogen into the core of a bicomponent jet. The fiber spun with 0.5 g/min polymer per hole at 1500m/min showed a 0.3 birefringence suggesting full orientation. Even at 1g/min/hole full orientation appeared to be achieved at only 1900m/min spin speed.
Questioners suggested that the birefringence measurements were affected by the hollowness, and that the fiber was not really fully oriented. Why did the hollow fiber have such a low tenacity and high extension? Ms Ortiz said that both solid and hollow fiber made on their machine had similar extensions, so the hollow result was not unusual. Why had nitrogen, rather than air been used to fill the core? She agreed air could be used in future.
Yanbo Lui of TANDEC gave Christine Sun’s paper. TANDEC made a 50/50 nylon/PP meltblown web and hydroentangled it. The fibers broke up into shorter fibers without splitting. The main conclusion was that basis weight reduces with increasing water pressure (due presumably to fiber loss), and that the adhesion between the two polymers in the bicomponent was too high.
Dr Gisela Buschle-Diller of Auburn University defined nanofibers as fibers with a diameter below 0.5 microns. When electrospun such fibers are invisible and when spun onto a black background it takes some time before the web begins to appear.
Polymers suitable for electrospinning should melt without decomposition or dissolve in a solvent and reform on solvent evaporation. In this study PVA, polycarbonate, biopolyester, poly-hydroxybutyrate, and poly(DL-lactide-co-glycolide) were spun through 5-25 Kv electric fields to a collector 5-25 cms away. Both voltage and distance were varied in 5 unit intervals.
Increasing the voltage gave finer fibers while increasing the nozzle-to-collector distance gave coarser ones. Melt-spinning polymers gave better results than solvent-spinning. Attempts to make a bicomponent sheath-core nanofiber resulted in side-by-side bicomponency.
Testing electrospun nanofibers proved difficult. However the coarser PLG fibers (1.8 micron or 0.03 denier/fil) gave a yield stress of 12 g/denier at a yield strain of 16%.
In response to questions, they have yet to try hot-drawing the nanofibers to improve strengths.
Peter Tsai of the University of Tennessee defined nanofibers as having diameters below 0.1 microns. Electrospun nanofibers can carry the spinning charge and Dr Tsai wanted to know what happened if electrically different polymers were spun, both separately and together. A variety of polar and non-polar polymers had therefore been spun and their charge-retention abilities had been monitored to see if useful electrets could be obtained.
No charge was retained on polar polymers but the non-polar materials had a potential of above 100volts for over 100 hours after production. Negatively charged polystyrene appeared best, retaining 250 volts for 250 hours.
The spread of fibers from the nozzle depended on fiber diameter, nylon (78nm) giving a 24” wide pattern while polyurethane (656nm) gave an 8” pattern. If 3 nozzles were used to spin the same polymer onto the same collector, the central web was narrowed by the repulsion of the outer webs, and the webs could not be made to overlap into a single sheet. If however alternate nozzles were oppositely charges the webs attracted one another and mingled, but tended to form ropes.
If 2 layers of oppositely charged polystyrene were spun on top of one another, the charges are not destroyed. Each surface maintains its own charge.
Phil Gibson of the US Army Natick Soldier Center was hoping to develop electrospinning to produce fabrics and filters to protect against chemical and biological warfare agents. He pointed out that electrospinning was really very simple technology and illustrated a hand-held device powered by 2x9volt batteries which could spray fibers. He envisaged such devices being used to overspray battledress garments to protect against spores and contaminated micro-dust when required. The fibers would have to be very elastic (nylon was too brittle!) and polyurethane appeared to be favoured. Enzymes could be added to the polymer to give them a specific reactivity.
If the PU was sprayed onto a woven wire, the fibers bond where they contact the wire and the resulting patterning provides a rip-stop effect which increases the tear resistance. Pore size was not affected.
A questioner thought that spraying onto a net would give a suitably reinforced structure. Dr Gibson agreed and said that this was still to be done. The net could be a nonconductive polymer, because it could be wetted to provide temporary conductivity.
Why not use polyurethane membranes in battledress? Because nanofibers were much better for comfort. Their porosity allowed convective as well as diffusive heat and moisture transport.
Kristine Graham of Donaldson Co. Inc. said that nanofiber webs were impossible to handle and so had to be directly applied to the product. Donaldson already spray them directly onto commercial filter materials, and are now trying to attach them to wearable fabrics for the same reasons as Natick.
When 0.25 micron fibers are sprayed directly onto a nylon shell fabric, the nanofiber layer tends to be punctured by any loose fiber ends. So Donaldson sprayed them onto 0.6 and 1.0 oz/yd 2 spunbond. To protect this layer, a top layer of similar spunbond was added to create a nanofiber sandwich. However better results were obtained by laminating two of the nanofiber-coated spunbonds together – nanofibers inside.
They now tried to attach this laminate to a shell fabric carrying activated carbon. Here beetter results were obtained if the two nanofiber layers were not bonded together, but allowed to float freely.
Washing showed up a fundamental problem. The nanofibers filtered out the detergent particles and after washing the fibers were therefore more contaminated than before. The filter proved too good to clean, but this did demonstrate the potential for attaching particles to the surface of the nanofibers
To demonstrate active chemistry potential, polyoxymettalate, a mustard gas adsorbent was added to a solution of thermoplastic polyurethane. This worked, and in fact there was evidence that the nanofiber enhanced the chemicals reactivity rather than diminishing it, as expected.
Asked how “heavy” the nanofiber coats were, Ms Graham said they don’t try to measure this. It would certainly be less than 0.1lb/ft 2. Could the detergent carrying effect be used to carry other useful chemicals. Yes.
Kevin Kit of the University of Tennesse, working with nylon 6,6 in a Dupont-funded project, was trying to align nanofibers to allow some drawing-down and to understand their properties.
He was spinning the fibers onto the sharp edge of a rapidly rotating meat-slicer to obtain aligned-bundles of nanofibers to be removed and tested.
The meat slicer had a critical rotational speed defined as the polymer extrusion velocity, and when running faster than this it stretched the emerging fibers. The nylon fibers were spun from a 10% solution in formic acid under 15 or 30Kv applied charge onto the blade of the slicer positioned 10 cm away. This gave 0.3micron fibers at a 0.8gms/day production rate.
Samples were collected by spinning for 8 hours at 5 speeds, from the critical speed (233m/min) up to 2833 m/min. At the critical speed, orientation was measured at 0.5 (i.e. none) and at the highest speed 0.86 (high). Molecular orientation went from 0.1 to 0.3 and crystalline orientation remained constant at 0.3. With high voltage extrusion mechanical properties increased only at the highest speeds, peaking at 120mPa for a 10% strain.
At 15kV the extrusion speed is lower so more orientation results and the strength increases more progressively over the collection speed range.
Dmitry Luzhansky of Donaldson Co Inc. thought fibers needed to be less than 0.25 microns in diameter to earn the nanofiber title. Donaldson make 10,000m 2/day of nanofiber webs on the surface of filters e.g the self-cleaning air-filter of the Abrams M1A1 battletank used in Iraq, and the combustion gas filters on gas-turbine engines.
Their quality control testing involves:
Continuous on-line non-contact monitoring of aerosol filtration efficiency uses laser particle counters upstream and downstream of the web. Measurements are taken at 5 positions across the web and the results are computer-analysed and printed out within 1 second of measurement. This is adequate for up to 200 m/min line speed.
What level of efficiency is targeted? 98% removal of 0.8 micron particles.
Samira Farboodmanesh of Lowell University described how adding a disc-shaped electrode to the nozzle could change the pattern of fibers on the collector. She had electrospun a 10% solution of polyethylene oxide in ethanol from a normal “needle” nozzle and from the same nozzle through a hole in various charged metal discs. Without the disc, a 20 cm spread of web was observed on the collector. With a 10 cm disc at the same voltage as the nozzle, the spread dropped to 10 cm and the diameter of the formed-fiber decreased. She did not know why the diameter decreased, but a member of the audience later suggested that it might be due to the increasing field strength created by the focussing effect of the disc.
If the charged disc is tilted at an angle to the needle-nozzle, it directs the web in that direction. If a charged plate is arranged parallel to the stream of fibers between nozzle and collector then the plate attracts or repels the fibers depending on charge.
The fibers can also be collected as a yarn in the gap between two charged needles. If one of these needles is rotated a small sample of twisted nanofiber yarn can be collected.
Kevin White of Physical Sciences Inc. is developing a process to reinforce the gossamer-thin films intended for NASA’s solar-sails. When you plan to use solar photons to push a spacecraft up to speeds approaching that of light itself, there is, to say the least, a great need to reduce the mass of the sail that does the pushing. NASA’s calculations show that a film has to be less than 15 micron thick to stand a chance of working, and at this thickness it is too fragile to be packed into and unpacked from the shuttle. Physical Sciences Inc is therefore trying to electrospin bands of nanofiber onto a 7.5 micron colorless polyimide film to reinforce it.
Their new (and subject to a patent application so it could not be shown) focussing system sprays on 1 metre lengths of a 2mm wide band of nanofibers, and this is repeated every half-metre. The total application is 0.2gms/metre and this increases the tear resistance of the film from 1gm to 28 gms.
Peter Tsai of the University of Tennessee(TANDEC) reviewed the ways of creating charged webs under the headings:
The best approach, and so the one used at TANDEC was corona treatment, and this could increase the filtration efficiency of a 90gsm 2-micron meltblown from 77 to 99.995%. However the charge dissipated under ambient conditions and dissipated more rapidly as particles (NaCl or Oil) built on the surface, or as temperature increased. The dissipation rate on PP could be reduced by use of a novel – but unspecified – “charge additive”, while the oil-induced dissipation could be reduced by the use of a fluorocarbon treatment.
Oil particles increased the filter efficiency of uncharged PP or nylon meltblowns, but did not affect uncharged PU or glass media.
John Hagewood of Hills Inc defined nanofibers as fibers less than a micron in diameter and pointed out that the Hills Inc thin-plate technology used to spin islands-in-a-sea fibers could, after extraction of the sacrificial sea polymer, yield very fine fibers.
A melt-blowing head with 100 holes per inch making IIAS bico at low throughput yields a spread of fiber diameters with a modal diameter in the 0.7 to 0.9 micron range.
In response to a question Mr Hagewood said they were now developing a 200 hole/inch head to allow higher productivity.
Rory Wolf of Enercon Industries Corporation explained how plasma enhanced chemical vapour deposition (PECVD) can now be used at atmospheric pressure to add high performance coatings to a variety of nonwovens, films and papers.
Successes claimed included:
Dr Todd Bullions of Virginia Polytechnic Institute and State University was looking for ways of using the 2-4 billion lbs of feathers produced by the US poultry industry. The cleaned and short-cut feathers had insufficient cohesion to be wet-laid on there own so Kraft pulp and kenaf had been evaluated as carrier fibers to get the feathers into sheet-form where their low density and good insulation properties could be utilised. Pulp proved to be the best carrier, but a kenaf enthusiast in the audience felt that the kenaf used here may have been a bad sample. Dr Bullions was advised to try again with a new one.
Dr D V Parikh, of USDA’s Southern Regional Research Centre described laboratory experiments involving the carboxymethylation of woven cotton gauze followed by treatment with silver nitrate to create an antimicrobial burn dressing. The silver nitrate was shown to be effective against staph. aureus, klebsiella pneumoniae and aspergillus flavus. However, if the silver nitrate concentration rose above 0.5%, e.g. by drying out, the burn dressing could burn you, so it had to be wetted regularly.
Thursday, 25 September 2003
Source: INDA INTC Baltimore Sept 2003