Recipient Organization
AUBURN UNIVERSITY
108 M. WHITE SMITH HALL
AUBURN,AL 36849
Performing Department
School of Forestry and Wildlife Sciences
Non Technical Summary
Over the past three years we have received United Soybean Board funding to study the possibility of including soy flour in MDI (methylene diphenyl diisocyanate) resin for manufacturing wood panels, principally oriented strand board (OSB). We have taken the approach such that 12% of MDI can be substituted with soy flour without any deterioration of performance (Cheng et al. 2019). Pilot studies run in 2018-2019 at 2 commercial plants (identity to remain confidential at their request) have confirmed these results and we are currently working with 2 more companies. The product lines of interest for these 4 companies include oriented strand board (OSB), particleboard, medium density fiberboard, and laminate columns.When OSB is used as an example, savings of up to $1 million can be realized per mill because soy flour is 67% cheaper than MDI (solids basis). This estimate was computed for mills making commodity OSB with 2% resin by weight. Specialty mills that manufacture premium OSB use 4% resin, and the savings will be proportionately higher. However, a mixing step (soy flour with MDI) has to be added and this step will create inherent risk due to increased viscosity. Likewise, industry is demanding at least 20% soy flour substitution before they will consider commercialization. The wet performance, such as edge swell, cannot be met at 12% soy substitution, primarily because we do not understand the interactions of wood, water, soy flour and MDI, especially at the glue line.?The objective of this proposal is to understand these fundamentals, develop conditions for meeting the 20% soy substitution requirement for a family of wood composites and run full scale mill trials if we can obtain external funds to demonstrate viability. In that vain, we have already submitted a proposal (pending) to AFRI NIFA to receive funds for OSB and other proposals will be written for MDF, particleboard, and columns. Other wood composite products may also be pursued if an industrial partner steps license our patent (Via et al. 2019). If successful, we will lower the cost of OSB, make it more sustainable, and increase the industrial applications of soy.
Animal Health Component
50%
Research Effort Categories
Basic
(N/A)
Applied
50%
Developmental
50%
Goals / Objectives
The proposed research will attempt to understand the interaction among soy flour, MDI, and water during hot pressing wood strands. We have been able to substitute about 12% soy flour in MDI resin but deterioration in wet properties have limited our ability to reach a higher substitution level. An understanding of this fundamental interaction is needed if we are to increase our substitution level to the >20% target requested by industry.This research will also strive to substitute MDI with soy flour so as to a) reduce cost b) maintain or improve performanceand c) increase sustainability
Project Methods
Water and soy protein will be profiled in wood as follows. Samples will be drilled from the board at approximately 0.5-inch intervals. The water content of the particles will be determined by standard Karl Fischer titrations. Determination of soy residues will be much more difficult because the soy flour will have residues from hydrolyzed and thermally degraded material and will also be to MDI. However, its Kjeldahl nitrogen content will be much higher than that of wood (< 0.01%, Skonieczna et al. 2014), so the presence of soy-derived material should be easily distinguishable from background signals. We will also measure reflectance IR spectra of the particles. The water band is broad and easily visible, although small changes may be difficult to distinguish. The bands originating from the soy products should be easily distinguishable from the background wood, especially since the soy will be covering the wood surface. We will assign the bands to various functional groups and then use them to assign the location of soy-derived samples on the board. We have extensive experience in quantitative IR spectroscopy, having developed a standard method for automatic spectral subtraction (Banerjee and Li 1991), now commercialized by Thermo Fisher as a part of their Spectral ID package.If, as we expect, the soy flour and water profiles board overlap, then more complete mixing of soy flour and MDI should alleviate or solve the problem as will be discussed But what if they do not overlap? This is unlikely because it is the presence of soy that degrades properties and it is difficult to conceive of a scenario where the soy and water would not be associated. Nevertheless, should this prove to be the case, then we will reduce the moisture content of the incoming flakes, e.g. about 7% to about 5%. This is not the preferred path because one of the advantages of MDI is that it can tolerate higher moisture flakes. Also, the higher moisture increases dryer throughput. In this situation, the economics of using the cheaper soy will need to be balanced against the need to dry the flakes to a slightly lower moisture content. The profiles will determine whether we adopt the better mixing or the water reduction strategy or a combination of the two.It would be convenient if we could measure the water distribution board without having to drill out samples. Our preliminary work on infrared thermography may offer a solution. Thermograms were taken with a high-resolution IR camera of boards immediately after press opening. The central 62,400 pixels were averaged from several boards. The temperature for the control boards averaged 171±4 oC, whereas the boards with 10% soy flour substitution were at 161±3 oC. In all cases the temperatures of the control boards were higher than those of the soy containing boards, so the difference is well outside statistical error. If the soy retained water, then the water in the board interior would be inhibited from moving to and moving heat from the board as it escaped; the surface would then be cooler than that of a corresponding control board, as observed. Correlating the temperature difference to the board moisture content would be very useful in field trials because the effect of soy substitution can be immediately demonstrated in pilot and field work. In earlier work we have imaged the drying of veneer to identify conditions that lead to press blows and have also quantified surface roughness of boards with high resolution thermography (Fike et al. 2004, Abedi and Banerjee 2007).We will collect 1D N-CPMAS NMR spectra over a temperature range of 20 oC to 70 oC in 10-degree intervals using 15N and 1H p/2 pulses of 6 ms and averaging of 1600 transients per free induction decay. To suppress proton-proton dipolar interactions, we will employ a proton evolution period of 0-16 BLEW-12 cycles of 72 ms each.Optimum spraying of this mixture onto flakes will be the next challenge. Conrad et al. (2003) and Smith (2003) have shown that fracture toughness is proportional to resin drop diameter, and inversely proportional to droplet spacing. At small droplet sizes, failure occurs at the interface. At larger droplet diameters cohesive failure occurs within the wood substrate with some failure in the interphase region. The optimum droplet diameter was between 290 µm and 405 µm, with a droplet spacing of less than 300 µm.We have developed a simple technique for imaging droplet size by spraying the resin on poster paper from a distance of 55 cm. The paper was then imaged by ImageJ software and the droplet diameter averaged from 400 drops (Cheng et al. 2019). We have also shown that the viscosity of MDI/soy flour mixtures is dramatically affected by temperature as shown in Figure 13. As discussed earlier, at 25oC the CO2 bubbles released from the reaction of MDI with soy flour and/or water are trapped within the predominantly MDI bubble, raising its viscosity. At higher temperatures the bubbles are able to escape and the viscosity falls (Cheng et al. 2019). Several mills heat their resin to 40oC, so these temperatures are in a practical range.The addition of soy flour to MDI resin (at 40 oC) affects its droplet size upon spraying as shown below. All the values are statistically similar, implying that the soy flour should not affect the spraying operation. Atomizers present in an industrial process will reduce the droplet size further, so any effect of soy flour on droplet size will be further attenuated. However, some nozzle clogging was observed at 20% soy flour substitution; the nominally low droplet value of 0.35 mm is a consequence of the larger particles being retained in the nozzle. Spraying at a slightly higher temperature should presumably reduce nozzle clogging. We will determine the best combination of soy flour substitution, soy particle size and temperature to get the resin size distribution closest to that of MDI alone. This optimum will be used to prepare boards for the measurements described below.Board testing procedures are straightforward and are standard at the Forest Products Center at Auburn (Cheng et al. 2019). A Zwick/Roell (Z010) materials testing machine will be utilized to test shear, lap shear, MOE, and MOR using loading rates and temperatures recommended by the appropriate standards. All samples will be conditioned to 22 ºC and 55% relative humidity unless a specific ASTM or other critical standard differs in their requirement.Wood flake composites will be made using standard procedures (Via 2013). The wood composite will be tested for modulus of rupture, modulus of elasticity, internal bond, and dimensional stability. A 6-cycle moisture exposure with vacuum pressure test has been suggested by our industrial partner as being the limiting factor that needs to be tested to be commercially viable (ASTM D5456 and D2718) in which, the new adhesive system needs to perform equivalent or better than the control panels and that from the industrial process. Finally, the optimum conditions developed in the study will be trialed in the field, with our OSB industrial partner. A pilot-scale press will be used initially. This press has been calibrated to full-scale operations and the results correlate well to full-scale operations. Full-scale mill trials will be run just before a scheduled maintenance shutdown to minimize downtime if practical problems (such as sticking to machine surfaces) should arise. The only problem we see during scale-up is that the mixing of soy flour with MDI will need industrial scale equipment. However, several inline powder liquid mixers have been developed for the food industry that we think can be leveraged to wood composite companies.?