INCORPORATION OF LIGNIN IN POLYURETHANE PRODUCTS

Abstract

Composition comprising lignin dispersed in a polyisocyanate wherein the (d90) mean particle size of the dispersed lignin is less than 5 m, preferably less than 2 m, more preferably less than 1 m and use of said composition in various polyurethane applications, in particular structural wood adhesives.

Claims

1. A composition comprising lignin dispersed in a polyisocyanate characterised in that the (d90) mean particle size of the dispersed lignin is less than 5 m.

2. The composition according to claim 1, wherein at least 50% of the dispersed lignin particles have a particle size of at least 500 nm.

3. The composition according to claim 1 wherein the lignin is free of ash content, sulphur content, and/or ion content.

4. The composition according to claim 3, wherein the ash content of the lignin is below 1 wt %.

5. The composition according to claim 3 wherein the suphur content of the lignin is below 1 wt %.

6. The composition according to claim 3, wherein the content of each of sodium, potassium, calcium and magnesium in the lignin is below 0.03 wt %.

7. The composition according to claim 1 wherein the lignin is an organosolv lignin or a Kraft lignin.

8. The composition according to claim 1 wherein the lignin has a moisture content of less than 10 wt %.

9. The composition according to claim 1 wherein the polyisocyanate is an aromatic polyisocyanate.

10. The composition according to claim 1 wherein the amount of lignin is between 1 and 25 wt %.

11. A process for producing a composition as defined in claim 1 using a three roll mill.

12. An isocyanate-terminated prepolymer, wherein the prepolymer is the reaction product of a stoichiometric excess of a polyisocyanate with a composition comprising lignin dispersed in an isocyanate-reactive compound wherein the (d90) mean particle size of the dispersed lignin is less than 5 m.

13. (canceled)

14. (canceled)

15. (canceled)

Description

EXAMPLE 1

[0090] Lap Shear Properties (Glue Line Thickness of 0.1 mm) when Dispersing Lignin Directly in the Isocyanate-Based Resin

[0091] Adhesive preparation: 20 wt % of dried Organosolv lignin (Alcell, Moisture Content 1 wt %) was dispersed into Suprasec 2144, an isocyanate-based resin of NCO value 15% available from Huntsman, in the following manner. A pre-weighted amount of lignin was mixed into the required amount of resin using a spatula until all particles were wetted. The mixture was stirred under nitrogen using a Heidolph mixer fitted with a Cowles blade at 3000 rpm for 30 minutes. Subsequently the 20 wt %-lignin-containing resin was processed by a three roll mill (model 80E from EXAKT) in three passes. For the three passes, the speed of the rollers was kept constant at 600 rpm, while the gap distance between the rollers was gradually decreased: for the first gap, from 60 m to 30 m and until 15 m, and for the second gap from 20 m to 10 m and until 5 m. The processed resin containing 20 wt % lignin was then further diluted to 10 wt %, 5 wt % and 1 wt %, respectively, using a Heidolph mixer fitted with a Teflon blade at 1500 rpm for 15 minutes under nitrogen. The isocyanate-based resins containing various lignin concentrations were stored in sealed glass bottles under nitrogen at room temperature.

[0092] Quality of the obtained resins with various lignin concentrations was evaluated using optical microscopy, viscosity and particle size measurements. An indication of the particle size was obtained by studying the dispersion using a Zeiss Jenavert Interphako microscope equipped with a DeltaPix camera. The viscosity was measured using a Brookfield R/S-CPS-P2+ rheometer fitted with a cone spindle (C25-2) for reference resin, or a plate spindle (P50) (with a gap of 85 microns) for lignin-containing resin. The measurements were performed at 25 C. and at a controlled shear stress. The stress was increased from 0 to 350 Pa in 1 minute, then kept constant at 350 Pa for 1 minute and subsequently decreased again to 0 Pa in 1 minute. Casson regression was applied to analyse the data. The viscosity of the resin was defined from the second step at a constant shear stress of 350 Pa.

[0093] The results are presented in Table 1 below.

[0094] Preparation of lap joints: Lap joints were prepared using tangential cut beech or pine wood (150205 mm). Wood was pre-conditioned in a climate chamber (22 C., 55% relative humidity) for at least one week, in order to obtain a moisture content of 100.5% and 140.5% respectively. Moisture content was measured using a Sartorius moisture balance. Lap joints were made with an overlap of 2020 mm and a resin loading of 500 g/m.sup.2 (0.2 g of resin) or 125 g/m.sup.2 (0.05 g of resin) applied on one substrate face by means of a brush. Before applying the glue, the wood surfaces were sanded with abrasive paper (P100) and dust removed. Resin was degassed in a SpeedMixer DAC 400.1 V-DP (2.5 minutes, 2500 rpm, 100% vacuum). Two series of 6 lap joints were usually prepared per lignin concentration for each condition. The lap joints were then pressed between 2 metal plates using spacers of 9.000.05 mm, resulting in a compression of 11-12%, and cured in a climate chamber (22 C., 55% relative humidity). After 24 hours, pressure was released and the lap joints further cured for 7 days under the same climate conditions.

[0095] Lap shear test: Lap joints were tested at 22 C. according to a modified method of the EN302-1 standard. Lap shear strength was measured using an Instron 5566 Universal Testing Machine and a crosshead speed of 50 mm/min. The extent of wood failure was assessed after breaking the sample. Two series of 6 lap joints per lignin concentration and condition were usually tested (total 12 samples), obtaining an average result of 12 values.

[0096] Results: Shear strength measured for beech lap joints prepared with 10 wt % Organosolv lignin-containing isocyanate-based resin (nominal glue line thickness of 0.1 mm) was found to be 10.4 MPa whereas for lap joints prepared with the lignin-free resin, a lap shear strength of 5.6 MPa was measured (improvement of 86%). Additionally wood failure was found to be increased from 0 to 45% by the incorporation of 10 wt % Organosolv lignin.

[0097] The moisture content of the beech wood was 10% and the resin loading was 500 g/m.sup.2. Independently of the wood species (pine or beech), resin loading (500 or 125 g/m.sup.2) or wood moisture content (10 or 14%), incorporation of lignin resulted in improved lap shear strength. Results are presented in Table 2 below.

TABLE-US-00002 TABLE 2 Beech wood MC 10%-500 g/m.sup.2 resin Pine MC 10%-500 g/m.sup.2 resin Lignin Stress @ max load Wood Stress @ max load Wood loading (MPa) failure % (MPa) failure % (wt %) Average St Dev (%) improvement Average St Dev (%) improvement 0 5.6 0.1 0 7.4 0.6 40 1 7.1 0.6 10 27 7.7 0.6 65 4 5 9.5 0.1 40 70 8.6 0.8 75 16 10 10.4 1.8 45 86 9.1 0.7 65 23 20 9.4 0.7 90 68 10.0 0.7 85 35 Beech wood MC 10%-125 g/m.sup.2 resin Beech wood MC 14%-500 g/m.sup.2 resin Lignin Stress @ max load Wood Stress @ max load Wood loading (MPa) failure % (MPa) failure % (wt %) Average St Dev (%) improvement Average St Dev (%) improvement 0 4.7 1.0 0 7.0 1.0 0 1 8.1 1.5 60 72 11.0 1.4 50 57 5 7.5 1.5 60 60 10.2 0.6 85 46 10 8.8 1.1 90 87 11.2 1.3 90 60 20

EXAMPLE 2

[0098] Gap Filling Properties (Glue Line Thickness of 1 mm) when Dispersing Lignin Directly in the Isocyanate-Based Resin

[0099] Adhesive preparation: same as in Example 1

[0100] Preparation of lap joints: Lap joints were prepared using tangentially faced beech or pine planks (1501105 mm). Wood was pre-conditioned in a climate chamber (22 C., 55% relative humidity) for at least one week, in order to obtain a moisture content of 100.5%. Moisture content was measured using a Sartorius moisture balance. In half of the planks, a 140 mm long, 10 mm width and 1 mm deep groove (2 mm from the edges) was carved (see FIG. 1). The wood surface was sanded with abrasive paper (P400) and dust removed. In case no lignin was used, 2 wt % of fumed silica (CAB-O-SIL TS720) was dispersed in the resin (using a Cowles blade at 1500 rpm for 10 minutes) as viscosity builder. Resin was degassed in a SpeedMixer DAC 400.1 V-DP (2.5 minutes, 2500 rpm, 100% vacuum). A second piece of wood with similar dimension but without a groove was then glued to the first by filling the groove with resin. The second piece was placed at 2 mm from the groove, as shown in FIG. 2. The billet was then pressed between 2 metal plates using a controlled gap of 9.500.05 mm, resulting in a compression of 9-10%, and cured in a climate chamber (22 C. and 55% relative humidity). After 24 hours, the pressure was released and the billet further cured in the climate chamber (same parameters) for 7 days. The billet was then cut along the length into 5 strips of 20 mm width (edges of the billet excluded), obtaining 5 gap filling lap joints (see FIG. 2). 2 series of 5 lap joints were usually prepared per lignin concentration.

[0101] Lap shear test: Lap joints were tested at 22 C. according to a modified method of the EN302-1 standard. Lap shear strength was measured using an Instron 5566 Universal Testing Machine and a crosshead speed of 50 mm/min. The extent of wood failure was assessed after breaking the sample. Two series of 5 lap joints per lignin concentration were usually tested (total 10 samples), obtaining an average result of 10 values.

[0102] Results: Shear strength measured for beech lap joints prepared with 20 wt % Organosolv lignin-containing isocyanate-based resin (nominal glue line thickness of 1 mm) was found to be 6.8 MPa whereas for lap joints prepared with the lignin-free resin a strength of 4.4 MPa was measured (improvement of 55%). Additionally wood failure was found to be increased from below 5% to 15% by the incorporation of 20 wt % Organosolv lignin. A minimum amount of 10 wt % lignin is required for this particular formulation in order to improve shear strength of lap joints with a 1 mm nominal glue line thickness. The results are presented in Table 3 below.

TABLE-US-00003 TABLE 3 Lignin loading Stress @ max load (MPa) Wood failure % (wt %) Average St Dev (%) improvement 0 4.4 0.3 0 1 4.0 0.4 0 9 5 4.2 0.4 0 5 10 5.8 0.3 5 32 20 6.8 1.0 15 55

EXAMPLE 3

[0103] Delamination Resistance of Mini-Glulams when Dispersing Lignin Directly in the Isocyanate-Based Resin

[0104] Adhesive preparation: same as in Example 1

[0105] Preparation of mini-glulams: Wood assemblies were made of 3 tangentially faced beech wood layers (2 glue lines). Each wood layer had a dimension of 2207020 mm. Wood was pre-conditioned in a climate chamber (22 C., 55% relative humidity) for at least one week, in order to obtain a moisture content of 100.5%. Moisture content was measured using a Sartorius moisture balance. Before gluing, the wood surfaces were sanded with abrasive paper (P400) and dust removed. A resin loading of 250 g/m.sup.2 was applied on 2 of the 3 pieces (top face only) using a brush. The wood layers were oriented to get 2 layers with the growth rings in the same direction, and the bottom layer with the grow rings in the opposite direction (see FIG. 3). The assembly was then pressed in a laboratory Schwabenthan press at room temperature for 24 hours using a specific pressure of 2.2 bars. After the pressure was released, the wood assembly was allowed to further cure in a climate chamber (22 C., 55% relative humidity) for at least 7 days.

[0106] After cutting of the edges of the wood assembly (from 1 to 2 cm), 4 mini-glulams of 405060 mm were cut out. Two wood assemblies were usually prepared per lignin concentration, resulting in 8 samples.

[0107] Delamination test: Mini-glulams were tested according to a modified method of the ASTM standard D2259, in a 3-cycle sequence described below:

TABLE-US-00004 1. The mini-glulams were immersed in water at 22 C. in a vacuum oven. 2. A vacuum was drawn and held for 5 minutes. 1st 3. The vacuum was released and the samples were left in water in the cycle {open oversize brace} vacuum oven for 1 hour. 4. Number 2 and 3 were repeated. 5. The mini-glulams were dried in an oven at 65 C. for 22 hours. 2nd 6. The mini-glulams were immersed in boiling water for 40 minutes. cycle {open oversize brace} 7. The mini-glulams were dried in an oven at 65 C. for 22 hours. 8. The mini-glulams were immersed in water at 22 C. in a vacuum oven. 9. A vacuum was drawn and held for 5 minutes. 3rd 10. The vacuum was released and the samples were left in water in the cycle {open oversize brace} vacuum oven for 1 hour. 11. Number 9 and 10 were repeated. 12. The mini-glulams were dried in an oven at 65 C. for 22 hours.

[0108] After each cycle, the two glue lines of each mini-glulam were visually assessed for signs of delamination. Delamination is expressed as percentage of the total glue line length:

[00001]$D=\frac{l_1}{l_2}100$

[0109] Where D =the delamination in percent [0110] 1.sub.1=the total delamination length in mm [0111] 1.sub.2=the total length of the glue line in mm

[0112] Two series of 4 mini-glulams were usually prepared and tested per lignin concentration. For each of them, 2 percentages of delamination were calculated (1 delamination percentage per glue line2 glue lines per mini-glulam), obtaining an average result of 16 values.

[0113] Results: Delamination percentage measured for beech mini-glulams prepared with 20 wt % Organosols lignin-containing isocyanate resin was found to be 13.5% whereas for mini-glulams prepared with the lignin-free resin a delamination percentage close to 100% was measured. The results are presented in Table 4 below.

TABLE-US-00005 TABLE 4 Lignin loading Delamination (%) (wt %) Average St Dev 0 94.2 11.5 1 100.0 0.0 5 50.0 0.0 10 68.7 2.5 20 13.5 8.2

EXAMPLE 4

[0114] Lap Shear Properties (Glue Line Thickness of 0.1 mm) when Dispersing Lignin in the Isocyanate-Reactive (Polyol) Segment of Resin before Prepolymerisation

[0115] Adhesive preparation: 20 wt % of dried Organosols lignin (Alcell, MC 1%) was dispersed into a mixture of Polyol A (functionality 2, MW 2000, 100% PO) and Polyol B (functionality 3.5, MW 3500, EO-tip of 15%) (50/50 wt ratio), polyol segment of the resin, in the following manner. A pre-weighted amount of lignin was mixed into the required amount of polyol mixture using a spatula until all particles were wetted. The mixture was stirred under nitrogen using a Heidolph mixer fitted with a Cowles blade at 3000 rpm for 30 minutes. Subsequently the mixture was processed by a three roll mill in six passes. For the first two passes, the speed of the rollers was kept constant at 200 rpm, while the gap distance between the rollers was gradually decreased: for the first gap, from 90 m to 75 m, and for the second gap from 30 m to 25 m. For the next three passes, the speed of the rollers was kept constant at 250 rpm, while the gap distance between the rollers was gradually further decreased: for the first gap, from 60 m to 45 m and until 30 m, and for the second gap from 20 m to 15 m and until 10 m. The same parameters as for the fifth pass were used for the last and sixth pass.

[0116] The quality of the 20wt %-lignin-containing polyol mixture was evaluated using optical microscopy and particle size was determined to be less than 5m (see FIG. 4).

[0117] Prior to pre-polymerisation, the 20 wt %- lignin-containing polyol mixture (after processing by a three roll mill) was dried under vacuum at room temperature (final MC 0.3%). The lignin-containing prepolymer was prepared according to the standard procedure. The required quantity of isocyanate (a uretonimine based isocyanate) (+10 ppm thionyl chloride to prevent trimerisation) was heated up at 80 C. in a glass flask under nitrogen flow and with continuous stirring (using a Teflon blade at 250 rpm). Then the specified amount of the 20 wt %-lignin-containing polyol mixture was added. Temperature of the reaction should be maintained at 80 C. After 2 hours of reaction, heating was stopped. After cooling down to room temperature (under nitrogen flow), a prepolymer containing 10 wt % lignin with a NCOv of 12% (referring to reaction between lignin and MDI) was obtained. This latter was then further diluted to 5 wt % and 1 wt % lignin using a Heidolph mixer fitted with a Teflon blade at 1500 rpm for 15 minutes under nitrogen. The prepolymers with various lignin concentrations were stored in sealed glass bottles under nitrogen at room temperature.

[0118] For 1 kg prepolymer was required: 525 g of the uretonimine based isocyanate and 475 g of 50/50 wt ratio mixture of Polyol A and Polyol B containing 20 wt % Alcell Organosolv lignin.

[0119] FTIR analysis of the prepolymer containing 10 wt % lignin has revealed formation of urethane bonds (1752 and 1725 cm.sup.1) resulting from the reaction between hydroxyl groups of lignin and MDI.

[0120] Quality of the obtained prepolymers with various lignin concentrations was evaluated using optical microscopy, viscosity and particle size measurements (see Table 5).

[0121] Preparation of lap joints: Lap joints were prepared using tangential cut beech wood (150205 mm). Wood was pre-conditioned in a climate chamber (22 C., 55% relative humidity) for at least one week, in order to obtain a moisture content of 100.5%. Moisture content was measured using a Sartorius moisture balance. Lap joints were made with an overlap of 2020 mm and a resin loading of 500 g/m.sup.2 (0.2 g of resin) applied on one substrate face by means of a brush. Before applying the glue, the wood surfaces were sanded with abrasive paper (P100) and dust removed. Resin was degassed in a SpeedMixer DAC 400.1 V-DP (2.5 minutes, 2500 rpm, 100% vacuum). Two series of 6 lap joints were usually prepared per lignin concentration. The lap joints were then pressed between 2 metal plates using spacers of 9.000.05 mm, resulting in a compression of 11-12%, and cured in a climate chamber (22 C., 55% relative humidity). After 24 hours, pressure was released and the lap joints were further cured for 7 days under the same climate conditions.

[0122] Lap shear test: same as in Example 1

[0123] Results: Shear strength measured for beech lap joints prepared with 5 wt % hardwood lignin-containing prepolymer (nominal glue line thickness of 0.1 mm) was found to be 9.2 MPa whereas for lap joints prepared with the lignin-free resin, a lap shear strength of 6.7 MPa was measured. Additionally wood failure was found to be increased from below 5% to 55% by the incorporation of 5 wt % hardwood lignin. The moisture content of the beech wood was 10% and the resin loading was 500 g/m.sup.2. When lignin was dispersed into the isocyanate-reactive (polyol) segment of resin (prior to pre-polymerisation), the optimal lignin loading was lower than when lignin was directly dispersed in the isocyanate-based resin (see Table 6).

TABLE-US-00006 TABLE 6 Lignin loading Stress @ max load (MPa) Wood failure % (wt %) Average St Dev (%) improvement 0 6.7 0.7 <5 1 7.4 0.9 15 10 5 9.2 0.9 55 37 10 7.9 1.1 55 18

EXAMPLE 5

[0124] Lap Shear Properties (Glue Line Thickness of 1 mm) when Dispersing Lignin Directly in the Isocyanate-Based Resin without 3-Roll Milling the DispersionLarger Particle Size

[0125] Adhesive preparation: 1, 5 and 10 wt % of dried Organosolv lignin (Alcell, MC 1%) was dispersed into an isocyanate-based resin (Suprasec 2144) in the following manner. A pre-weighted amount of lignin was mixed into the required amount of resin using a spatula until all particles were wetted. The mixture was stirred under nitrogen using a Heidolph mixer fitted with a Cowles blade at 3000 rpm for 30 minutes.

[0126] Quality of the obtained dispersion with various lignin concentrations was evaluated using optical microscopy, viscosity and particle size measurements.

[0127] The particle size of the lignin was significantly higher for dispersions prepared by Heidolph mixing only compared to 3-roll milling, and viscosity is significant lower for high lignin concentrations (Table 7).

[0128] Preparation of lap joints: see example 2

[0129] Lap shear test: see example 2

[0130] Results: Shear strength measured for pine lap joints prepared with 1, 5 and 10 wt % Organosolv lignin-containing isocyanate-based resin (nominal glue line thickness of 1 mm) was found to be on average respectively 2.2, 2.2 and 4.2 MPa, whereas for lap joints prepared with the lignin-free resin, an average lap shear strength of 1.5 MPa was measured (average improvement of respectively 39, 42 and over 100%). However a very wide spread in the results (>35% relative standard deviation) was obtained due to the low quality lignin dispersions (high particle size) whereas dispersion prepared using 3-roll milling show a relative standard deviation of <15% in the results. Strengths measured for 1 and 5 wt % lignin are in the worst case respectively 25 and 15% lower than the lignin-free system and in the best case both over 100% improved. When incorporating 10 wt % lignin, in all cases an improvement was observed (min. 37% and max over 300%).

[0131] The results are presented in Table 8.

TABLE-US-00007 TABLE 8 Stress @ max load (MPa) Improvement (%) St Dev av high low System Avg St Dev (%) low high ref ref ref S2144 1.6 0.1 7.7 1.4 1.7 S2144 + 1% Li 2.2 0.9 42.6 1.2 3.1 39 26 115 S2144 + 5% Li 2.2 0.8 35.6 1.4 3.0 42 15 109 S2144 + 10% Li 4.2 1.9 45.8 2.3 6.1 172 37 329

EXAMPLE 6

[0132] Lap Shear Strength (Glue Line Thickness of 0.1 mm) when Dispersing Lignin Directly in the Isocyanate-Based Resin Without 3-Roll Milling the DispersionLarger Particle Size

Adhesive Preparation:

[0133] 5 and 10 wt % of dried Organosolv lignin (Alcell, MC 1%) was dispersed into an isocyanate-based resin (Suprasec 2144) in the following manner. A pre-weighted amount of lignin was mixed into the required amount of resin using a spatula until all particles were wetted. The mixture was stirred under nitrogen using a Heidolph mixer fitted with a Cowles blade at 3000 rpm for 30 minutes. The dispersion was split into two. Subsequently, one part of the dispersion was processed by a three roll mill as described in Example 1.

[0134] Preparation of lap joints: see example 1

[0135] Lap shear test: see example 1

Results:

[0136] Quality of the obtained dispersion with various lignin concentrations was evaluated using a Grindometer. The particle size of the lignin was significantly higher for dispersions prepared by Heidolph mixing only compared to 3-roll milling (see Table 9 below). Lap shear results of beech lap joints with a glue line thickness of 0.1 mm (MC of 10%) and prepared with a resin loading of 250 g/m.sup.2 can be found in Table 9 below. Improvement in lap shear strength is much higher when dispersion is 3-roll milled. Also high percentages of wood failure are observed whereas for the dispersions prepared with Cowles blade only, no wood failure at all was obtained.

TABLE-US-00008 TABLE 9 Lignin Particle Stress @ max load Wood loading Dispersing Viscosity size (MPa) failure Improvement (wt %) method (Poise) (m) Average St Dev (%) (%) 0 19 5.4 0.5 0 5 CB 22 15-25 5.9 1.1 0 9 10 CB 24 15-25 5.0 0.8 0 7 0 19 5.4 0.5 0 5 CB + 3RM 49 <1 8.9 0.6 30 65 10 CB + 3RM 63 <1 9.2 0.7 72 70 Dispersing methods: CB = Cowles blade, CB + 3RM = Cowles blade + 3 roll-mill