Vegetable tanning agents can be divided into two groups, hydrolysable and condensed tannins. There are structural differences between the two types and by reacting hydrolysable vegetable tannins (gallotannins or ellagitannins) with metal salts, leathers exhibiting high shrinkage temperatures, to match those obtained with chrome tannage, can be achieved. This can be viewed as extending the multiple interactions between the polyphenol molecules and collagen by creating a crosslink but, more importantly, by reducing rotations within the protein matrix1. In this way, there is a positive effect on both the crystallinity and the effective cooperating unit, which leads to an increased hydrothermal stability.

The question of how the metal interacts with the vegetable tannin has led to a number of investigations. Slabbert2 has studied the combination of mimosa tannage and aluminium sulfate retannage. He suggested that the aluminium ions are fixed to the already bound flavanoid units of the, in this case, mimosa via complexation. This was based on the work of Sykes, Hancock and Orszulik3 who showed that aluminium complexes with mixed s-phenolate and carboxylate groups. Slabbert suggested the binding occurs as in figure 1. Kallenberger4 set out to contradict the notion that any covalent or co-ordinate covalent ties to the collagen are formed by the tannin/aluminium treatment. To demonstrate the lack of importance of the protein/metal interaction in producing a very heat stable leather, he used metals that have little or no conventional tanning power.

The metals were chosen for their multivalent nature in the belief that there was a need to link the tannin into polymers to achieve stability. The commercial stock from which the samples were taken was a wet vegetable tanned hide from an American tannery. At that time tanners used a blend of chestnut, wattle and quebracho. Most varied the mix depending upon price, so if quebracho was cheap that was increased while higher priced components were cut. Generally, wattle and quebracho were the principal tannins with a little chestnut added for colour5. Thus, the effects of the metal were not optimised. Some of the results are set out in table 1.

Although the vegetable tannins are not identified, the valence states of the metal ions chosen were not unequivocal and the offers used not stated, there are a number of general conclusions that can be drawn. Covington6 recast the results in terms of three interactions: metal collagen, metal veg tanned and metal veg tannin. These are shown in table 2.

The data shows that in most cases there is a powerful interaction due to the addition of the vegetable tanning agent to the system. Even with the manganese(II), which is not a known tanning agent as can be seen from the effect of metal alone on the collagen stability, the effect of adding the vegetable tanning agent is to increase the shrinkage temperature by 10°C. With other metals the effect is more dramatic. The most interesting results are those for metal vegetable tannin interaction. When the metal salts were used, the effect is the same for Fe(III), Ni(II) and Co(II), and smaller, although still marked, for manganese(II). The results do not necessarily show that Slabbert’s model can be discounted because although the metal vegetable tannin interaction is large, it is not, in general, greater than the metal-veg tanned collagen interaction. Hence, the results are not inconsistent with Slabbert’s postulate of a mixed crosslinking system as shown in figure 1.

In a review paper7, Covington suggests that the synergistic interaction between the polyphenol and the aluminium may arise from one of the following options:

1. Collagen-Al-Veg-Al-Collagen

2. Collagen-Veg-Al-Veg-Collagen

3. Collagen-Veg-Al-Collagen

He contends that since it is known that applying the metal salt before the vegetable tannin produces only moderate shrinkage temperature, the first and third options are unlikely. The metal is likely to crosslink the vegetable tannin, figure 2, and, in effect, the crosslinked polyphenol is itself crosslinked to form a matrix within the collagen matrix. In other words, the reaction appears to be dominated by polyphenol complexation. The collagen is then stabilised by a multiplicity of connected hydrogen bonds. This is not inconsistent with the recent work carried out by Song at the BSLT8.

To investigate the effects of metal on the hydrothermal stability and rate of shrinkage of vegetable leathers, various metals were used to retan the leathers tanned with myrabalams or sumach. Samples of the vegetable tannages were retanned in petri dishes with different metals using 1% metal on the weight of the sample of vegetable tanned leather. The metals used, together with the shrinkage temperatures recorded at different time intervals, are set out in table 3.

It has been shown9 that, except at low temperatures, and after a short period of induction, shrinking follows approximately an exponential function of time, expressed by the following equation lt = l0-l. exp(-kt)+ l. where lt = length of sample at time t, l0 = the initial length, and l• = length of sample at infinity, in this case when shrinkage had stopped. k = rate constant and t = time in seconds. A graph of ln (lt – l.) verses t should yield a straight line of slope k, the rate constant.

Some of the semi-metal tannages were subjected to measurement of the rate of reaction. This procedure followed that developed by Weir, and used the standard shrinkage apparatus, as set out in SLP 1810, with a 0.5cm graduated scale added to the arm holding the sample.

The leather was held at the previously determined shrinkage temperature in the apparatus in a water bath and allowed to shrink. The first result was noted after 10 seconds, to allow for relaxation of the sample, when the dial was reset to vertical. The readings were taken, noting the time the leather shrank past a graduation until no further measurable shrinkage noted. All the vegetable tannages were with mimosa unless stated. The results are set out in table 4. With myrabalams leather, retanned with titanium(III) or vanadium(IV), the rate of shrinkage is ten times slower than the other semi-metal leathers initially tanned with the mimosa. As the composition of myrabalams and mimosa is different, the former composed of ellagitannins, while the latter is made up of polymers of catechins, it seems the effect is due to a different interaction between the vegetable tannin and the metal, rather than the interaction between the metal veg complex and the collagen or the metal and the collagen.

This lends further weight to the idea that the polyphenol complexation is the important reaction. The results also show that the rate of shrinkage is independent of the type of metal, although the actual shrinkage temperature value is not.