Cooking Chemistry Transforms Proteins into High-Strength Adhesives

Subscriber access provided by YORK UNIV

Article

Cooking Chemistry Transforms Proteins into High Strength Adhesives Jessica K. Roman, and Jonathan J. Wilker J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b12150 • Publication Date (Web): 21 Dec 2018 Downloaded from http://pubs.acs.org on December 21, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Cooking Chemistry Transforms Proteins into High Strength Adhesives Jessica K. Román1ϯ and Jonathan J. Wilker1,2* 1

Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, IN 47907-2084, USA.

2

School of Materials Engineering, Purdue University, Neil Armstrong Hall of Engineering, 701 West Stadium Avenue, West Lafayette, IN 47907-2045, USA.

Abstract: In prior generations, proteins were taken from horses and other animals to make glues. Petroleum-derived polymers including epoxies and cyanoacrylates have since replaced proteins owing to improved performance. These modern materials come at a cost of toxicity as well as being derived from limited resources. Ideally, replacement adhesives will be made from benign, cheap, and renewable feedstocks. Such a transition to bio-based materials, however, will not occur until similar or improved performance can be achieved. We have discovered that coupling of proteins and sugars gives rise to strong adhesives. An unexpected connection was made between adhesion and Maillard chemistry, known to be at the heart of cooking foods. Cross-linked proteins bonded metal and wood with high strengths, in some cases showing forces exceeding those withstood by the substrates, themselves. Simple cooking chemistry may provide a route to future high-performance materials derived from low cost, environmentally benign components. Introduction Industrial adhesives are used on massive scales with, for example, 9 billion kilograms produced annually in the United States.1,2 Nearly all of these glues are petroleumderived, permanent in nature, and almost half off-gas carcinogenic formaldehyde.1,2 Wood composites such as plywood, particle board, medium density fiberboards, and floor laminates give rise to a significant fraction of our daily formaldehyde exposure, be it from houses or furniture.1 Consequently, we have a tremendous demand for new adhesives that are obtained sustainably, not releasing toxins, and of low cost. Proteins might provide one of the best sources of cheap, renewable, and non-toxic organics for adhesives. Early successes with adhesives, centuries ago, did include hide glue comprised of collagen obtained from animal connective tissue.3 Today, horses are marched to the glue factory less often. The twentieth century petrochemical boom has supplanted most use of bio-based adhesives in favor of synthetic systems.4 Cyanoacrylates, epoxies, and urethanes often exhibit high structural bond strengths not achievable previously with biomacromolecules, albeit with accompanying toxicity and even some cost concerns. Despite persistent problems, formaldehydeurea and formaldehyde-phenol glues still dominate within wood products given the low costs. Avoiding the drawbacks of these modern materials will only happen when we can match the performance of petroleum-based materials with cheap and renewable feedstocks such as proteins. On their own, proteins are not generally too sticky. However, marine biology has developed several clever strategies here. Mussels, barnacles, and oysters bind themselves to rocks quite well using protein-based attachment strategies.5-9 As we learn how marine

organisms stick, an emerging theme is the cross-linking of proteins. This chemistry is complex, far from understood well, and somewhat unusual in that crosslinking may be a product of post-translational modifications allowing oxidation reactions to then take place. By analogy to shellfish, cross-linking chemistry will be needed if we wish to make use of such organics available on large scales. Biomimetic design of materials is both exciting and a popular means of gaining access to adhesion. Simple polymers such as polystyrene can be transformed into high strength glues upon incorporation of mussel chemistry in the form of 3,4dihydroxyphenylalanine (DOPA).10 Such functional groups have also been appended to, for example, soy protein with excellent adhesion results.11 Yet biomimetics often include synthetic challenges, extensive modifications of proteins,12,13 increased costs, and may only take us so far. We still need affordable and simple chemistry that can be scaled up to millions of tons per year. While exploring other types of cross-linking, we noticed that simple reactions between proteins and ascorbic acid were creating materials with appreciable adhesive properties. We have now applied these reactions to biopolymers that can be sourced on large scales. When conditions were just right, high strength adhesion resulted. In some cases, performance was comparable to the commercial products in most need of replacement yet without requiring the high heat and pressure often required in composite material manufacturing. Delving into the origin of this adhesion discovered that one of the oldest known means of coupling proteins was at play. Reducing sugars can react with proteins to generate cross-links. We are, perhaps, most familiar with such chemistry when cooking and 1

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

baking food.14-17 Consider making bread and watching the dough become brown. This Maillard chemistry is a dominant process in all cooking, darkening foods via cross-linking reactions and releasing numerous flavorful compounds. By combining insights from shellfish, biomimetics, and cooking chemistry, we now have the opportunity to impact how we make the adhesives all around us.

shear testing on wood substrates was performed the same way with the exception of the amount of material deposited, 45 μL vs 7.5 μL. In most cases, the adherends were allowed to set for 30 minutes, after which time the weights were removed and the adherends allowed to cure at 37 °C for 23 hours followed by cooling for 30 minutes. Bulk adhesion was tested in shear using an Instron 5544 Material Testing System with a 2,000 N load cell and a crosshead speed of 2 mm min-1. The method used is a modification of the ASTM D1002 standard method. For each study, a data set of at least 10 trials was collected. The average of these data sets and errors at 90% confidence intervals are reported. A statistical difference of p < 0.05 from that of the control in each parameter optimization was considered significant.

Materials and Methods Cross-Linking Reactions. All protein + ascorbic acid reactions were completed in 50 mL Falcon tubes with protein dissolved in 3 mL of double distilled, deionized water. These tubes were then Parafilmed shut and shaken on their side at ~150 rpm in a New Brunswick Scientific incubator. Bovine serum albumin was purchased from Sigma Aldrich and soy protein isolate (92% protein) was received from MP Biomedicals. Protein Characterization. Amino acid analysis was carried out at the Molecular Structure Facility of the University of California, Davis. Hydrolysis was completed at 110 °C over 24 hours with the addition of 200 μL, 2 M HCl with 1% phenol. This facility utilizes Hitachi amino acid analyzers, which separate amino acid residues via ion-exchange chromatography followed by a ninhydrin reaction and detection system. Viscosity measurements were carried out using FungilabTM BS/U-Tube Capillary Viscometers (ASTM #4, 5, and 6). Samples for UV-vis and circular dichroism were removed from a reaction vessel at T = 0 with the controls made at the same concentration as their reaction counterpart. UV-vis analysis was carried out on a Varian Cary 100 Bio UV-vis spectrophotometer. Circular dichroism samples were further diluted to protein concentrations of 0.025 mg/mL and analyzed by a Jasco J-1500 spectrophotometer. Samples for FTIR analysis were removed after the optimal reaction times and lyophilized to dryness. These materials were then analyzed by a Thermo Nicolet 6700 FTIR/FT-Raman spectrophotometer. Adhesion Testing. All adhesion testing was carried out using aluminum and pine wood substrates. Sheets of aluminum, 6061-T6 (Farmer’s Copper), were cut into adherends (8.89 cm x 1.27 cm x 0.318 cm) and cleaned following the ASTM D2651-01 standard method. Common pine, purchased from a local hardware store, was cut to 8.89 cm x 1.27 cm x 1.27 cm, holes drilled for testing with rods, and used without any further surface modification. After adhesion testing, the strength was determined by dividing the maximum force at failure (Newtons) by the overlap area, determined by a Vernier caliper. Lap shear testing of protein samples on aluminum was done generally as follows. The protein solution (7.5 μL) was spread onto two adherends using a micropipette, overlapped (1.2 x 1.2 cm) in the single lap-shear arrangement, and pressure applied with 55 g weights. Lap

Page 2 of 10

Results and Discussion Two different systems were chosen to explore the transformation of proteins into high strength adhesives. Bovine serum albumin (BSA) is one of the cheapest proteins that can be purchased on large scales within a laboratory setting, providing a pure material with which to best understand the chemistry at hand. Soy protein isolate is less pure and not fully soluble, but is one of the lowest cost sources of organic biopolymers available today at pennies per kilogram in bulk, in addition to being biodegradable, even following crosslinking.18 Being under a dollar per kilogram may enable soy-based organics to be cost competitive with formaldehyde-based wood adhesives, which comprise nearly half of all industrial glues in use today.1,2 These proteins were subjected to several reaction conditions and adhesive testing in order to explore the possibility of generating high strength, protein-based bonding. Initial screening of conditions found suitable starting points to include protein concentrations of 0.05 grams/mL for BSA or 0.30 grams/mL for soy, 25 °C, pH = 7, and mixing times of either 1 or 7 days. These formulations were then used for initial bulk adhesive bonding of aluminum and pine (i.e., wood) substrates (Figures S1, S2). After placement of the protein solutions onto one substrate, a second substrate was added to create a lap shear joint. Typical cure conditions were then 30 minutes at room temperature followed by 23 hours at 37 °C and then 30 minutes at room temperature. Adhesive bonding was quantified by pulling the substrates apart until failure. The reducing sugar ascorbic acid (i.e., vitamin C) was combined with BSA and soy protein at varied molar ratios (protein: ascorbic acid). Figure 1(A, B) shows that pronounced influences upon adhesion were found. Bonding near ~1 MPa was achievable with BSA on aluminum and soy protein on wood. Without the addition of any ascorbic acid, adhesion was in the 0-0.1 MPa range. Ascorbic acid enhanced adhesion, but only up to a point, after which performance decreased. When ascorbic acid was in excess over protein (i.e., ratios of 1:10, 1:50, 1:100), the 2


Page 3 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


solutions became increasingly viscous, possibly due to excessive protein cross-linking. In addition to creating cohesive bonding at the expense of surface adhesive bonding, this viscosity may also have reduced the ability of solutions to penetrate into the wood substrates and create mechanical interlocking. Solution pH also influenced adhesion, although values both higher and lower than pH = 7 decreased performance (Figure S3), possibly due to diminished solubility, with

prior reports illustrating decreased soy protein solubility at lower pH.19-20 Reaction times of ascorbic acid plus protein, prior to applying the mixtures between substrates, altered bonding significantly, seen in Figure 1(C, D). For BSA a large jump in wood adhesion came at 14 days. Longer than two weeks showed less strong joints, likely a result of over cross-linking. With soy protein, earlier time points up to 3 days had appreciable bonding, but further reaction time decreased solubility as

A

B

C

D

E

F

Figure 1. The effects of ascorbic acid concentration (moles of protein to ascorbic acid, A & B) at 25 °C, reaction time (C & D), and protein concentration (E & F) on the lap shear adhesion of bovine serum albumin (A, C, E) and soy protein isolate (B, D, F) on aluminum and wood substrates. Reaction time was tested with the optimal molar ratio of protein to ascorbic acid and concentration tests used the optimal reaction time. The reaction time and concentration experiments used the optimal reaction temperatures of 25 °C (BSA-aluminum, soy-aluminum, and soy-wood) and 37 °C (BSAwood). Error bars represent 90% confidence intervals and asterisks identify data sets that were statistically (p ≤ 0.05) different from the control value in each experiment (1:0 protein : ascorbic acid, 0 days, and 0.05 g/mL).

3



well as adhesion. Achieving a proper viscosity, neither too thin nor too thick, is one of many factors in developing an adhesive, with increased viscosity caused by factors beyond concentration including intramolecular forces from cross-linking. Here, in Figure 1(E, F), concentration was varied from a water-like 0.05 grams/mL up to a more paste-like 0.2 grams/mL for soy protein and 0.6 grams/mL for BSA. High bonding at ~2 MPa was noted for BSA at 0.3 grams/mL and wood substrates. Interestingly, when bonding wood, soy protein decreased strengths at higher concentrations, possibly related to increased viscosity and diminished penetration into the surface. On aluminum, by contrast, a middle concentration proved more optimal. Metals ions have been implicated in cross-linking within several contexts including proteins involved in Alzheimer’s disease21 and marine bioadhesives.7 Here we assessed the effect of added Na+, Cu2+, Fe2+, Zn2+, and Fe3+ salts. Large changes were not observed for most ions (data not shown) although Fe3+ did increase BSA adhesion on wood by ~0.5 MPa. For soy protein, Fe3+ increased aluminum bonding by ~0.5 MPa (Figure S4). Harsh oxidants such as dichromate [(Cr2O7)2-] are often used to cross-link proteins when, for example, tanning leather.22 Here we examined potential reaction with sodium periodate (NaIO4), a strong oxidant without the established carcinogenicity of chromium. Changes in the material failure mode on the aluminum substrates were found with this oxidant. Higher NaIO4 concentrations resulted in more adhesive failure, evidenced by, after the completion of measurements, protein debonded from the substrates. Cohesive failure was exemplified, after pulling apart two substrates, by an even distribution of glue on each substrate and was common in this study up until the point of adding an oxidant. The BSA and periodate bonding increased on wood by ~1 MPa up to ~3.8 MPa (Figure S5A). Although periodate decreased soy protein adhesion on wood, there was an increase on aluminum up to ~0.7 MPa (Figure S5B). Curing of adhesives is often a tradeoff between time and elevated temperature. Both loss of solvent and induction of cross-linking chemistry can be at play. Here, cure times of 1-24 hours and temperatures of room (25 °C), physiological (37 °C), and high (60, 95, 150 °C) were examined. Potentially relevant are the denaturation temperatures of BSA (50, 80 °C) and soy protein (76, 97 °C).23,24 Pressure sensitive (e.g., tapes) and hot melt adhesives are often designed to manage glass transition and melting temperatures according to targeted performance characteristics.25,26 For these experiments the adhesives were placed between substrates, held together for only one minute, cured under the relevant conditions, and cooled for only 3 minutes prior to mechanical testing. Cooling time was kept at the minimum required for the substrates and glue to cool ensuring no additional cure time at room temperature, which could have skewed the parameter

Page 4 of 10

A

B

C

D

Figure 2. Testing the adhesion strength of protein adhesives. (a) Solid wood substrates shown in the testing system, which pulls the two substrates apart with pins. (b) Broken wood substrate in which the metal pin ripped through the wood rather than breaking the adhesive bond. (c) A close-up photograph of the broken substrate in frame (b). (d) Plywood substrates in which the joint created from a bovine serum albumin adhesive broke one adherend due to high adhesion strength. optimization. In the case of soy protein on aluminum (Figure S6C) trends of higher adhesion with longer cures or elevated temperatures was fairly clear. Trends for analogous systems were not as obvious (Figure S6), although several general observations could be made. After pulling apart the substrates during testing, the adhesives remained tacky and not dry with short cure times or low temperatures. Conversely, longer cures and higher heat yielded drier, more brittle materials often with higher bond values. Being quite inexpensive, soy protein isolate contains other components such as ash (4.1%) and fats (0.8%) in addition to protein (92%).27 Solubility and, subsequently, solution viscosity increased with formulations made at higher temperatures, likely a result of known changes to protein dispersibility indices with temperature.28 The subsequent decrease in solubility may be responsible for lower, yet still substantial, soy bonding relative to BSA. Table 1 provides specific conditions for the BSA and soy protein adhesives found to be optimal on aluminum and wood. The viscosity of glues can play a large role in abilities to bond, especially with porous substrates. Soy protein 4


Page 5 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1: Comparison of the final conditions to provide maximum bonding for Maillard protein adhesives. Optimal BSA conditions

Optimal soy conditions

Aluminum

Wood

Aluminum

Wood

moles protein : ascorbic acid

10:1

10:1

1:1

1:1

reaction pH

7.0

7.0

7.0

7.0

reaction temperature

25°C

37°C

25°C

25°C

reaction time

7 days

14 days

1 day

1 day

protein concentration

0.30 g/mL

0.30 g/mL

0.10 g/mL

0.05 g/mL

metal ions (protein : metal)

No metal

Fe3+ (10:1)

No metal

Fe3+ (10:1)

cross-linker (protein : IO4-)

No IO4-

10:1

100:1

No IO4-

cure time

3 hrs.

3 hrs.

12 hrs.

12 hrs.

cure temperature

95 °C

95 °C

95 °C

95 °C

Table 2. Adhesive performance of several protein-based commercial glues tested under the final, optimized conditions of both bovine serum albumin and soy protein isolate on aluminum and wood substrates. Deviations represents 90% confidence intervals and asterisks identify data sets that were statistically (p ≤ 0.05) different from the Maillard protein adhesives. Performed under optimized BSA conditions

Performed under optimized soy conditions

Material

Class

Aluminum adhesion (MPa)

Wood adhesion (MPa)

Aluminum adhesion (MPa)

Wood adhesion (MPa)

NW139C

protein

0*

0*

0*

0*

Superset

protein

0.1 ± 0.2 *

0.2 ± 0.1 *

0*

0.2 ± 0.1 *

CM261C

protein

0.2 ± 0.1 *

0.6 ± 0.2 *

0*

0.6 ± 0.2 *

Prolia

protein

1.1 ± 0.3 *

1.9 ± 0.7 *

0.5 ± 0.2 *

0*

Sealmaster P30L

polysaccharide

0.3 ± 0.4 *

2±1*

0.4 ± 0.1 *

4±1*

AP240

polysaccharide

0.2 ± 0.2 *

3±2

0.2 ± 0.1 *

1.7 ± 0.6

Gorilla Glue

synthetic

3±1

2.8 ± 0.3 *

2±1

1.4 ± 0.8

BSA

protein

2.8 ± 0.7

4.0 ± 0.5

N/A

N/A

Soy

protein

N/A

N/A

1.5 ± 0.2

2.0 ± 0.3

5



solutions were measured for room temperature viscosities and found, at optimized conditions, to be ~50 centistokes (cSt, wood) and ~60 cSt (aluminum). The BSA solutions were significantly more viscous, at ~150 cSt (aluminum) and ~500 cSt (wood). For comparison the viscosity of water measured at ~1 cSt. Along with concentration changes intramolecular forces, including cross-linking, may be at play with the highly viscous adhesive formulations. Replacing petroleum-based adhesives with renewable feedstocks is likely to occur only when performance and cost can be comparable. Methods for generating adhesives from proteins are being discovered constantly. One visible example is the incorporation, via a multi-step

Page 6 of 10

synthesis, of marine mussel adhesive chemistry into soy protein.11 Animal-derived hide glues, casein adhesives from milk, and the biomedical BioGlue all have proteins for the main components, although performance of these systems tends to not be in league with more modern petroleum-derived systems. How did the performance of these new materials compare to analogous commercially available products? Conditions vary from study to study, thereby making correlations between literature reports not terribly insightful. Thus side-by-side testing under identical conditions was carried out here. Several protein-based commercial glues were examined (LD Davis Industries NW139C, Superset, and CM261C as well as Cargill Prolia). Further comparisons were made

A

B

C

D

Figure 3: Characterization of protein adhesives. (a) Ultraviolet-visible spectra observing the absorbance change in bovine serum albumin after reaction with ascorbic acid. (b) Circular dichroism spectra of bovine serum albumin after reaction with various ascorbic acid concentrations. Ratios indicate bovine serum albumin : ascorbic levels. (c) Infrared spectra of bovine serum albumin, pre- and post-reaction, dashed lines mark 3292, 2960, 1271 and 1109 cm-1. (d) Amino acid analysis before and after the Maillard reaction, which proceeded for 7 days prior to analysis. Error bars represent 90% confidence intervals.

6


Page 7 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


α-helix and β-sheet structure were noted, however only at particularly high ratios such as 1:100 equivalents of BSA : ascorbic acid.30 With such excesses of ascorbic acid not correlating to where the best performance was found, protein denaturation did not appear to be a key factor in obtaining the strongest adhesion. Although the optimized adhesive conditions suggested little to no secondary structure changes, Fourier transform infrared spectra (Figure 3c) indicated possible protein-ascorbic acid conjugation. Changes in the absorption band located at 3250-3400 cm-1 could be attributed to the -NH2 and -OH stretch vibrations,31 a potential modification of amino groups in that region. Increased absorption between 2800-3000 cm-1 and 1050-1250 cm-1 may be changes to the carboxylic acids in those regions.32 Perhaps most insightful was amino acid analysis of BSA performed before versus after reaction with ascorbic acid (Figure 3d). Of the 20 amino acids, lysine was the one showing the most significant decrease, thus indicating that this amine-containing sidechain was the primary protein site of cross-linking.

to starch-based glues (Grain Processing Corporation Sealmaster P30L and LD Davis Industries AP240) as well as a modern, petroleum-derived polyurethane (Gorilla Glue). Table 2 shows how the cross-linked proteins performed relative to these products. For aluminum bonding, the highest measured strength was for BSA at ~2.8 MPa, with performance similar to Gorilla Glue. The highest lap shear bond strengths measured were when joining wood substrates. One commercial product (Sealmaster) and the new BSA protein system both yielded strengths of ~4 MPa. Figure 2 shows that the joint created between wood pieces could be stronger than the wood, itself. When trying to pull apart bonded substrates, the wood failed while the adhesive joint remained intact (Figures 2B, 2C). Likewise, adhesion was strong enough with plywood that substrate failure was seen during experiments (Figure 2d). Initial testing of the optimized soy and BSA protein adhesives with polyvinyl chloride (PVC) and polytetrafluoroethylene (Teflon) substrates resulted in weak bonding (

Cooking Chemistry Transforms Proteins into High-Strength Adhesives

Recommend Documents