Quo vadit chromium?20 August 2007
Introduction With the recently developed 'link-lock'1 theory the 'link' part of the reaction concerns the primary reactant, the main component that reacts directly with the collagen. All tanning reactions exhibit this part of the mechanism, which may involve any or all types of reaction: hydrogen bonding, hydrophobic bonding, electrostatic interaction and covalent bonding. The effects are twofold:
First, the structure of the supramolecular water2 around the triple helices is modified, thereby changing the ability of the protein to spread into the interstices between the triple helices, when the helical structure begins to unravel as it denatures under hydrothermal conditions. All primary reactions produce the same outcome, with only a small variation: the observed shrinkage temperature of single component tannages is typically 75-85°C, although some may confer even less stability because of the nature of the chemical change. The actual value achieved depends on the effect of the reaction on the entropy of activation3, which controls the rate of shrinking, reflected by the conventionally measured shrinkage temperature. Part of the modification is the nature of the primary interaction: the weaker the interaction, the lower the hydrothermal stability, but the upper limit to the hydrothermal stability is defined by the strongest interaction, covalent bonding. The degree of structure or additional bonding introduced by the tanning reaction has only a second order effect on the hydrothermal stability: regardless of the type of structure imposed by this reaction, the hydrothermal stability is limited. Second, the presence of the tanning agent modifies the shrinking reaction. The ease with which the triple helix can unravel into the region between triple helices depends on the resistance offered by the intervening chemical species. The mechanism of the shrinking reaction is independent of the chemistry of tanning1, so the effect of the single tanning species is to hinder the outcome of unravelling. This view is consistent with the observation that any tanning reaction from a single component, which can only cause the linking effect, will confer the same moderate hydrothermal stability increase. Qualitatively, the effect of the primary reactant is to interfere with the unravelling process, as the collagen converts from the helical structure into the random coil structure, hence the similarity of outcome. In addition, the interaction between the tanning species and the supramolecular water modifies the shrinking process, giving rise to the observed small variations in the moderate hydrothermal stability resulting from the single component reaction. The outcome of the linking step is independent of the ability of the tanning agent to crosslink. From the work of Das Gupta4, mono-functional oxazolidine and bi-functional oxazolidine confer the same shrinkage temperature, about 80°C. The 'lock' part of the mechanism refers to the action of a second component in the tannage. The second reaction can take several forms: * The second tanning reaction may be independent of the first tanning reaction. For example, in chromium III retanning after vegetable tanning, regardless of the polyphenol type, hydrolysable or condensed, the chromium III fixation reaction occurs with the carboxyls of the collagen and not with the phenolic hydroxyls5. Therefore, the characteristics of the outcomes of the reactions are individual, rather than combined. * The second reagent may interact with the first, but still have no effect on the outcome of the combined process. The second reagent may react with the first reagent, creating a new tanning reaction; in other words the interaction is synergistic. This refers to the situation when the observed effect on the hydrothermal stability is greater than the sum of the contributions: ΔTsobs > ΔTs1 + ΔTs2 * where ΔTs refers to the effect (rise) in shrinkage temperature conferred by the tannage. Note, if the sides of the equation are equal, the effects of the tanning reactions are independent and additive (to some extent), as described above. If the observed shrinkage temperature change is less than the sum of the contributions, the reactions are antagonistic, ie one interferes with the other, to decrease its effect. The difference between the observed shrinkage temperature and the calculated sum of the contributions is indicative of the magnitude of the effect in creating a new tanning species and hence of the new reaction on the stability of the leather. * The second reagent may additionally react both with the first reagent and the collagen; in this way it can link the matrix more firmly to the collagen. This happens in the combination reaction of condensed polyphenols and oxazolidine9, when the dual reactions confer a positive effect to the tanning outcome. If there is synergistic interaction between two components of the tanning reaction, the primary reactants are locked together by the second reagent, effectively changing the individual reactions into a concerted, single interaction. The consequence of forming a single chemical entity around the triple helices is to make it much more difficult for the collagen to collapse into the interstices. This is observed as high hydrothermal stability. It is critical that the reaction should take this form: merely filling the interstices with unlinked molecules, as happens in full vegetable tanning, cannot prevent molecular movement in the collagen, as the triple helices unravel. There are two observations in the shrinking transition that must be incorporated into the ideas of the mechanism of tanning. It has been demonstrated that the shrinking reaction is independent of the tanning reaction, because the energy is the same. The shrinking transition does not involve the breaking of the artificial links between the tanning agent and the collagen. It has, therefore, been postulated that the shrinking reaction involves the breaking of hydrogen bonds within the collagen structure. The new view of the structure of collagen, proposed here, which includes the presence of structural water in the supramolecular matrix, must play some part in the reaction. Indeed it has been shown that the water associated with the collagen structure is practically independent of the tanning chemistry: the amount is the same, about 23%, whether the collagen is stabilised with chromium III, formaldehyde or high levels of vegetable tannin, hydrolysable or condensed11. It can be further postulated that there must be some of this structural water remaining in the matrix modified by the tanning chemistry. Therefore, shrinking involves the unravelling of the triple helix, aided by the breakdown of the residual structural water associated with the triple helix. It has been demonstrated that the shrinkage temperature of collagen depends on its moisture content: this has been explained in terms of the close approach of the triple helices, as water is removed: the effect is to prevent the unravelling process, since there is decreased space for this to happen12. Of course, this is already known in leather technology, because it underpins the requirement for the shrinkage temperature to be conventionally measured under conditions of water saturation, when the state of the sample is simply defined and easily achieved in practice, so the shrinkage temperature is at its lowest. This can be understood in terms of the ease of displacement of the intervening molecular species during shrinking. Therefore, the idea that hydrothermal stability depends on the chemical situation within the collagen structure is consistent with the known role that water plays in the structure of collagen and hence also plays in modified collagen. The important conclusions to be drawn from this analysis are as follows: Single component tannages are restricted by the thermodynamics of the shrinking reaction to confer only moderate shrinkage temperature. High shrinkage temperature can only be achieved by tanning with at least two conventional components. High shrinkage temperature is achievable if there is strong primary chemical bonding between the tanning agent and the collagen and if the primary tanning agent is locked in place by a secondary tanning agent. The chemistry of the reactions is less important than conforming to the requirements. Therefore, high shrinkage temperature is achievable by many more routes than is currently known. The arguments for the first and second conclusions have been presented above. The third conclusion can be rationalised by considering tannages that do conform and do not conform to this requirement. If the locked matrix is strongly bound to the triple helix, the ease with which the matrix can be displaced during unravelling is greatly affected, making it more difficult for the reaction to occur: this is observed as high hydrothermal stability, when more energy must be applied, to break down the water-matrix structure. Alternatively, the formation of a matrix that is weakly bound to the triple helix or not bound at all infers a greater role for the structural water. For example, tanning with aluminium III salts can be compared with chromium III tanning: there are superficial similarities, but the outcome is quite different, because aluminium III salts alone are incapable of conferring high hydrothermal stability. However, it has been shown that the chemical environment of the aluminium III species is not changed by the shrinking reaction10. The possible interactions between aluminium III and collagen are modelled in Figure 2. The (basic) aquo ion can interact electrostatically via a water ligand or a complex might be formed, which would be more electrovalent than covalent. Since it is known that aluminium III does not form stable complexes, it may be surmised that the former case is more likely. Therefore the bonding between the collagen and the matrix could break down hydrothermally, but the environment of the aluminium nucleus would not be substantially altered. Therefore, it can be concluded that the aluminium-based matrix involves water that can break down, to allow shrinking, ie the electrostatic interaction with collagen carboxyls is sufficiently distant to allow this to happen, but the aluminium III nucleus experiences no change in its magnetic field. Alternatively, covalent complexation between collagen carboxyls and chromium III is a direct interaction, which cannot break down under the conditions of shrinking. The future of chrome tanning The environmental impact of chromium III is low: as a reagent, basic chromium salts are safe to use industrially and can be managed, to the extent that discharges can routinely be as low as a few parts per million. Therefore, it is clear that the industry should be able to meet all the future requirements of environmental impact. Consequently, we can reasonably assume that the future of tanning will include a major role for chrome. Nevertheless, the technology can be improved: efficiency of use can be improved, as can the outcome of the reaction, in terms of the performance of the leather (shrinkage temperature) and the effectiveness of the reaction (shrinkage temperature rise per unit bound chrome)13. It might be a cause for curiosity as to how this reaction, the best known high hydrothermal stability tannage, fits into the link-lock mechanism. It has been shown14 that chrome tanning is not, in fact a solo tannage; it is a combination reaction, comprising the chromium III molecular ions and the counterion, where the counterion is not acting as a ligand in the chrome complex15, illustrated in Figure 3. Chromium chloride or perchlorate confer only moderate hydrothermal stability to collagen, but when sulfate or any other locking ion is introduced, the familiar high shrinkage temperature is obtained16,17. Therefore, the process might be reconsidered in this light, that is, as a two stage/step reaction. Since there is an environmental problem with sulfate (as a solubiliser of concrete), the use of an alternative salt and another locking counterion would be advantageous. It is useful to compare the conventional view of chrome tanning, the assumption of collagen sidechain crosslinking, and the new proposition, the matrix model. If the conventional view is correct, the tanning reaction should be independent of the nature of the counterion. However, that is clearly not the case: tanning is only moderately effective for chromium III chloride or perchlorate, it is effective in the presence of sulfate, but its effect is almost maximised in the presence of pyromellitate, 1,2,4,5-tetracarboxy benzene, if it acts as a counterion rather than a ligand18. Therefore, the matrix model explains the observations, but the traditional view does not. The uniqueness of chrome tanning is not that it constitutes a single tanning agent, but that the linking and locking components of the mechanism may be applied at the same time. The tanner's approach to chrome tanning can be enhanced and extended by including consideration of the role of the solvent and the nature of the chrome tanning complex. It has been argued14 that the reaction between a solute and a heterogeneous substrate can be expressed in the form of stepwise reactions, as follows: 1. Transfer from the solvent into the substrate: (solute) solvated + (substrate) solvated = solute-substrate) solvated The equilibrium depends on the relative affinity of the solute for the solvent and the affinity of the solute for the substrate: the more hydrophobic the solute, the greater its relative affinity for the more hydrophobic environment of the substrate. Since this is a mixed phase step, its contribution to the reaction kinetics is to influence the partitioning of the solute between the phases, thereby facilitating the availability of the solute for the reaction within the substrate, but the kinetics of reaction are not affected, since the mechanism involves the chemical reactions of the bonding interaction19,20. Therefore, the effect of the transfer step is only an apparent change to the kinetics of reaction. Consequently, we can distinguish a differentiation between the rate of uptake of the solute and the rate of fixation. This can be understood more clearly by using the model of non aqueous solvent tanning of the type suggested by Wei21. Here, tanning of conventionally pickled, wet pelt is conducted in paraffin, as the tumbling and heating medium. Chrome tanning powder is added: the salt is insoluble in the paraffin, but is rapidly solubilised in the wet substrate. Although the uptake is highly efficient and effective, 100% uptake in a matter of minutes, the fixation is a different reaction, which proceeds much more slowly. Despite this example being more concerned with solubility than transfer, it can be regarded as an extreme example of the general case. The role of the solvent in chrome tanning has already been exploited: Rohm and Haas marketed Chromesaver A30, ethanolamine hydrochloride HOCH2CH2NH3+Cl-, as a pickling auxiliary22, which changes the dielectric properties of the solution, even at low concentrations. It is not immediately obvious that the solvent is changed by adding this solid product, but it becomes clearer if the action were to be reconstituted by mixing the organic liquid ethanolamine with the pickle solution and readjusting the pH to the required value. Strictly, in either case, the effect is to create a mixed aqueous organic solvent. It is generally true that tanners usually do not use water in their processing. Consider the typical conditions in solution: often the process liquors are concentrated solutions of salts of different types, so the solvent properties are significantly different to water. In the thermodynamic sense, dilute solutions are defined as <10-4 molal, above that concentration the solutes modify the water properties by creating an electric field. The higher the electrolyte concentration, the more the solvent functions as a charged medium, with charge-charge interactions between solvent and solute. An example of the effect is the technology of adding salt to the bath when using reactive dyes. Here, the charged nature of the brine means there is less affinity of the solvent for the dye, thereby driving the dye into the more hydrophobic environment of the substrate. In this way, the balance of the relative rates of hydrolysis and uptake are altered in the tanner's favour. The effect of the species in solution should be considered when analysing any reaction and the same kind of analysis can be used to create improved or new outcomes. 2. Electrostatic interaction between the solute and the substrate All reactions of this type involve an initial electrostatic interaction, whether it is a simple charge-charge attraction between the reactants or a wider aspect of electrostatics, hydrophobic and hydrogen bonding. The contribution of this step extends into the initial step: the more highly the solute is charged, the greater the potential attraction into an oppositely charged substrate. Even though charge on the solute will contribute towards the solvating effect of aqueous solvent and therefore affect the apparent rate of reaction, the overall reaction of complexation may not be dependent on a rate determining step of primary interaction by electrostatic attraction. An example of this step operating is given by vegetable tanning23, when the transfer is facilitated by the hydrophobic nature of the polyphenols, giving rise to the conventional view that the initial interaction with the collagen is via hydrophobic bonding. The bonding then becomes converted to hydrogen bonding, as the ultimate electrostatic interaction. In the case of carboxylate complexation with chromium III in solution, it has been shown that the rate of reaction is independent of the carboxylate compound, since the initial electrostatic interaction is fast, but the rate determining step is proton loss19. The charge on the complex will affect the interaction between the carboxylate and the charged complex, but this is a fast rate, not rate determining. Therefore, the charge on the complex is probably more important in determining the hydrophilic-hydrophobic properties and hence the initial attraction of the substrate for the solute. It is important to understand that all carboxylate complexation reactions run at the same rate, including chelation, so the rate of masking is the same as the rate of tanning24. This clearly has implications for the masked status of the chrome species at any time in the tanning process, depending on the chemical conditions before and during the reaction14. The role of masking is further discussed below. 3. Covalent reaction Initial charge based interaction may be followed by covalent reaction, if the chemistry allows, as it does in chromium III complexation or tanning with condensed polyphenols. The solvent-solute interaction is sometimes characterised by the value of the hydrophilic-hydrophobic balance (HHB), alternatively designated hydrophilic-lipophilic balance (HLB), quantitatively measured by the chromatographic movement of the solute in a range of solvents with different dielectric properties. This is the parameter that controls the transfer reaction, probably most familiar to tanners with regard to dye properties. For example, the Sellaset dyes have matched HHB properties, so their rates of reaction are the same and they appear to react as if the mixture was a single dye. In the case of chromium III, the solvent interaction can be controlled by the ligand field, referred to in the jargon as masking. This is familiar technology: for example, the use of phthalate to make the chrome more reactive. Contrary to the popular belief that the dibasic phthalate ion makes the chrome more reactive by crosslinking molecular ions, it chelates a chromium atom and makes the molecular ion hydrophobic, due the presence of the benzene ring in the ligand field. Indeed, if the reaction is crosslinking, the effect on the reactivity of the chrome would be a second order effect1. Because the availability of reaction sites for complexation is at the ends of the chrome species, the trans sites, the reactivity would be practically unaffected by the masking reaction and the effect on the reactivity would be less dependent on the influence of the crosslinking agent on the HHB value of the complex. The technology of masking must be reviewed, because there is much misunderstanding regarding its effects on chrome tanning. Introducing a ligand into the chrome complex can be considered to have two competing outcomes: * The reactivity of the chrome species to complexation with collagen is assumed to be reduced, because there is a statistical reduction in the number of available reaction sites, illustrated in Figure 4. In the masking reaction, as in the tanning reaction, substitution occurs at the trans positions of the octahedral complex, the positions in the plane of the dihydroxy bridges, shown in Figure 415. * The reactivity of the masked chrome species can be enhanced, because the complex becomes more hydrophobic by the introduction of the carboxylate into the ligand sphere and all carboxylates are more hydrophobic than an aquo ligand - at least those commonly used in the art. The effects of masking with formate and other masking salts, which can react in different ways, are shown in Table II. Here the masking ratio is 1 mole per mole Cr2O3: the effect of the number of equivalents of carboxyl per mole should be taken into account. In each case, the effect of the masking on the reactivity of the chrome species is designated, based on the known chemistry of the masking agent and the observation of the relative extent of reaction. The designation of 'hydrophobic masking' indicates the influence of the masking agent on the properties of the complex and hence the element of the reaction mechanism that is affected: increasing the hydrophobicity of the complex encourages transfer from aqueous solution. The designation of 'complexation masking' indicates that the net effect is to modify the ability of the complex to react further, ie this conforms more to the conventional view of masking. Comparing the effects of formate and oxalate masking in Figure 5, the notional structures indicate that formate is likely to be less able to interact with water than an aquo ligand, thereby conferring a small degree of hydrophobicity. On the other hand, oxalate has the potential for effective hydrogen bonding, because it offers only carbonyl groups for interaction with the solvent, so the net effect is a reduction in the number of available reaction sites and hence reduced affinity for collagen carboxyls, driven by the chelating mechanism. The designation of 'molecular size' indicates the influence of crosslinking, to polymerise the molecular ions. This in itself will increase the rate of chrome fixation, since, even if the rate of complexation is unchanged, the rate of chrome uptake is increased. Table 1 illustrates the following important points: a) If we assume that chrome content is a reflection of the overall reaction rate or reactivity, there is no indication that masking with formate at this level reduces the reactivity of the chrome salt; indeed, there is an increase in reactivity. This runs contrary to conventional thinking about the masking effect of formate. It must be recognised that the relative effects of masking are dependent on the masking ratio. Those apparently contradictory effects are: the reduction in reactivity by using up reaction sites and the increase in reactivity by increasing the hydrophobicity. Taking the example of formate, as the simplest and commonest masking agent, at higher masking ratios, [RCO2]:[Cr2O3]>1, the reactivity reducing effect will begin to dominate, typically resulting in a reduction in tanning efficiency26. At lower masking ratios, which is more typically related to industrial practice (the masking ratio of one mole of formate per mole of chrome oxide corresponds to a formic acid offer of 0.6% if the chrome offer is 2% Cr2O3), the outcome is an enhancement of chrome reactivity. The effect of incorporating formate has the same hydrophobic effect as masking with malonate; the latter has a more predictable impact on the complex properties, because of the presence of the methylene group, but the outcome is similar for formate and malonate masking. b) In recognising the role of masking ratio, the combination of the chemical outcome with the kinetics of the process must also be taken into account. Unless the chrome is masked by conducting the reaction separately, prior to starting the tanning process, the masking ratio will necessarily be low at the beginning of the chrome tanning reaction and then increase as the tanning reaction proceeds: the progress and change of masking ration will depend on the chrome fixation kinetics, controlled by the pH profile, and on the offers of chrome and masking agent. The influence of masking ratio on chrome reactivity is modelled in Figure 6. The assumed effect is to reduce the reactivity by eliminating reaction sites, but the actual effect is to include some enhancement of reactivity by changing the HHB value of the ion. In Figure 7 the relative effects of those two contributing parameters will depend strongly on the chemistry of the masking agent. The role of masking in practical tanning depends on the tanner understanding the nature of the actual effect, even only qualitatively. c) The extent of increasing the reactivity of the masked chrome is predictable based on the evident chemical nature of the masking ligand, ie the CH=CH π-bond of the maleate can interact with water more than the saturated linked methylene groups of succinate, which are in turn less hydrophobic than the large benzene ring of phthalate. d) Crosslinking masking depends on the calculated chelate ring size exceeding seven: for aliphatic dicarboxylates, it starts with glutarate and the effect of the four methylene groups in the chain of adipate is clear in the reactivity. Therefore, all crosslinking masking agents are likely to increase the reactivity of chrome: the effect is not concerned with the inherent reactivity towards complex formation, but is dependent on the enhancement of fixation rate by polymerisation. The role of masking in chrome tanning is an important feature, which can be technologically exploited. The use of specific masking agents, which gradually increase the astringency of the chrome species, by increasing their tendency to transfer from solution to substrate, is an aspect of the reaction that has not received scientific attention. Here, the requirement is to match the rate of diminishing concentration of chrome in solution with the rate of masking complexation, to maintain or increase the rate of chrome uptake. This type of reaction is already technologically exploited by the use of disodium phthalate, although the degree of hydrophobicity conferred by even a very low masking ratio can cause undesirably fast surface reaction. Other hydrophobic masking agents could be developed, to give a more controllable increase in astringency. It is already recognised that the chrome tanning reaction is controlled by the effect of pH on the reactivity of the collagen substrate: a second order effect is the increasing of the hydrophobicity of the chrome species by polymerisation. In addition, the rate of reaction is controlled by temperature. However, the efficiency of the process is limited by the role of the solvent, in retaining the reactant in solution by solvation: this is typically only countered by applying extreme conditions of pH, likely to create problems of surface fixation, causing staining and resistance to dyeing27. The role of the counterion in chrome tanning offers potential for change. It has been shown14,16,17 that the chrome tanning reaction is controlled by the particular counterion present. It was fortunate for the leather industry that chrome alum (potassium chromium III sulfate hydrate) was the most readily available salt for the original trials of tanning ability: sulfate ion is highly effective in creating a stable supramolecular matrix, because it is a structure maker in water1,13. The effect is very different if other salts are used, eg chloride or perchlorate, when the outcome is only moderate hydrothermal stability. However, even if these salts are used, the high hydrothermal stability can be acquired by treating the leather with another counterion. Since the effect is independent of a complexing reaction, the process of modifying the moderate tanning effect is fast. This opens up the chrome tanning reaction to modifications which exploit the separation of the link and lock reactions. The environmentally damaging sulfate ion might be replaced by other less damaging counterions, such as nitrate. Note, typical sulfur dioxide reduced chrome tanning powder contains about 50% sodium sulfate. The reactive counterions can then be applied; options include using the stoichiometric quantity of sulfate or organic anions. The counterion might be replaced with polymeric agents, including polyacrylates with the right steric properties. Conclusions The expression of the link-lock mechanism has made developments in tanning technology feasible. This new theory is a simpler and more powerful view of collagen stabilisation than the older model of direct crosslinking between adjacent sidechains; it is a more elegant view. It is no longer worth pursuing the single reagent alternative to chrome tanning, because it does not exist. Indeed, why should we seek an alternative to chrome tanning? It works well, it can be made to work even better and, anyway, by all reasonable judgements it causes little environmental impact. The continuation of developing tanning and leather technology depends on constant reappraisal of all aspects of the subject. This is the role of leather science. Conventional, received wisdom should not be relied upon without critically reviewing exactly what it means, what it contributes to processing and products and what the wider implications are for the practical tanner. It is important to recognise that the scrutiny of current technology will often identify inconsistencies and misunderstanding of principles: the technology may work, but the science may not. However, this is not always a bad thing, because it can lead to new thinking, new developments and more profitability in an environmentally sound, sustainable industry. This paper goes on to consider alternatives to chrome.