There are many different types of science. Some are loud and explosive, some are quiet and precise, and some are just kind of smelly. The MA project I'm working on this year definitely falls in that last category and I apologise to my fellow students for that.

Within some types of furniture metal decoration will be inlaid into wood. The most famous of this type is probably Boulle furniture, well known for its elaborate, twirly decorations and its unfortunate tendency to fall apart. To stick the metal onto the wood, animal glues, made from hide or bone, would be used. These types of adhesives are very effective when gluing organic materials, but don't tend to work very well on metal. The glue doesn't form strong bonds with the metal surface and this often causes the metal inlay to become loose or even get lost completely. In the trade there are many tips and tricks for improving the adhesion of metal to wood. Some heavily abrade the surface of the metal, making deep grooves or roughly sanding it. Others add other ingredients to the glue, such as beer or even urine.

The trick that I'm interested in, however, is rubbing garlic onto the metal surface. This technique is described as either etching the metal surface or changing some properties of the glue, such as viscosity or wettability. These changes should make the bond between the metal and the wood stronger so that the decoration may stay in place for at least a little bit longer. I want to find out if this technique works and if so, how it works.

I'm going about this in two steps. Firstly I will be rubbing lots of garlic onto lots of brass samples to see if I can create some etching. That is, after all, supposed to be one of the factors that make this technique work. Secondly I will be making some samples of wood glued to brass and I will pull them apart using a tensometer. Some of these samples will have no additions to them and others will be treated with the garlic. I can then compare these two groups and determine if the addition of garlic has improved the strength of the adhesive joint.

A single lap joint of one part wood and one part metal. The arrows indicate the direction of the force that will be put on these samples.

I've completed the first set of preliminary tests. A set of 6 samples was rubbed with garlic and the garlic juice was left to dry for about three minutes. Looking at the samples under the microscope seemed to show that the garlic was indeed etching the metal. The scratches in the surface that were made during polishing seemed to be getting deeper. Afterwards, however, when the surface was cleaned completely using acetone, the "etching" disappeared. What I thought might have been etching, turned out to simply be the residue of garlic reflecting the light differently. So the assumption that the garlic etches the metal surface appears to be wrong. The next step is to make absolutely sure that the garlic doesn't etch the surface by taking that testing to the extreme: rubbing the garlic on repeatedly until it can be said with certainty that etching will not happen.

Microscopic image of a brass surface. The metal on the right appears to be etched, but this is actually just caused by the residue of garlic juice.

And what then? If I do find out that there is no etching due to garlic juice, then I need to take that into account when I do the second part of my testing. Because if I find that the samples with the garlic added to them are stronger than those in the control group, then what is causing the difference? Does the garlic residue on the surface itself provide a key for the adhesive? Does it act as a primer for the animal glue to attach to?

It will take some more testing and sadly also some more smelly times in our analytical lab before I can answer those questions. But perhaps I can provide an update by the end of the summer and hopefully by then I won't be sick of garlic just yet.