Multi-Part Report

Corrosion at Cabo da Roca -6

Origin of Sulphur

The sixth part of a multi-part series


It has been my observation that, almost without exception, highly corrosive sea cliffs are associated with very high levels of sulphate, either coating the rock surface, or oozing from the groundwater. Further to this fact is the observation that, in almost all cases, an exceptional source of sulphur is easily accounted for by the presence of submarine volcanic activity. In the case of Cabo da Roca, however, whilst we have sulphate in abundance, its source is not altogether obvious.

I will argue that the sulphur source is the Tyrrhenian Sea of the Western Mediterranean. The same sulphur source that is likely responsible for the corrosive sea cliffs of Sardinia, Ibiza, Mallorca and Penyal d’Ifac.

What we know:

Let’s start by reviewing the evidence that I’ve covered in previous posts.

The Sintra volcanic complex, which is comprised mainly of granites and syenites, forms impressive, jagged sea cliffs that extend for several kilometres both north and south of Cabo da Roca. Syenite is a low-silica form of granite and looks very similar.

Unlike typical granitic intrusions with their massive, sparsely-jointed structures, these rocks are highly fractured with a marked NE/SW jointing. It is seems likely that they are highly permeable to ground water.

The sea cliffs predominantly face westward into the open Atlantic Ocean. They are sparsely vegetated for quite some distance back from the cliff top.

The rugged coastline immediately south of Cabo da Roca showing extensive vertical jointing that strikes SE/NW.

At many locations white crystalline deposits form where ground water seeps from cracks.

Sulphate efflorescence at Espinhaco.

Examination of this material shows it to be remarkably pure calcium sulphate.

It is slightly soluble in water as would be expected, and aside from this, the only freely soluble components are to be found in the presence of a small amount of sodium chloride, as illustrated below.

A further observation of note is the fact that sulphate is found in sections of rock that are significantly isolated from the mainland. Consider, for example, the case of Aroeira pictured below. The presence of sulphate on on the rocky ridge indicated by the arrow argues against the likelihood of an inland source.

This sharp rocky ridge at Aoeira is significantly detached from the land yet, drilling the rock at point indicated, reveals elevated levels of sulphate.

Wall-wash sampling reveals nothing unusual apart from the occasional elevated sulphate level. We see a buffered calcium bicarbonate system partially displacing marine sodium chloride. And, as is typical for other sea cliffs, the marine magnesium signal is often suppressed.

Electrolyte composition of wall-wash samples taken from various Cabo da Roca crags.

The elevation of calcium in the surface electrolyte mix is puzzling. It is unlikely to originate from seawater given the observed reversal in the magnesium/calcium ratio, and yet, this leaves us explaining how large amounts of calcium, especially large when you consider the quantities of calcium sulphate crystalising from the ground water, can be derived from granite/syenite without also elevating silica, aluminium and potassium. Yes, silica is a little elevated, but potassium is hovering at seawater levels, and aluminium is not detectable at all.

It has been my experience that other granite sea cliffs also exhibit marked calcium bicarbonate buffer systems and, given that these other examples are sulphate-free, non-corrosive crags, it seems likely that calcium accumulation is more a feature of the granite sea cliff environment than anything peculiar to Cabo da Roca.

While it is likely that the calcium is locally derived, the source of the sulphate is a wholly different matter. Certainly, seawater contains appreciable amounts of sulphate, but as I have commented elsewhere, for whatever reason, the levels are insufficient to support attack by sulphate reducing bacteria. Corrosive crags are always associated with sulphate levels much higher than can be provided by seawater alone.

If we examine the data we now have available to us, these being drawn from a large number of samples taken across the globe, we can draw some reasonably solid conclusions about a possible marine origin for the sulphate.

Firstly, we note that the chloride burden imposed by the presence of the sea does not extend much further inland than 1km. This finding is in keeping with the idea that the main carrier is the coarse aerosol fraction generated by the breaking of waves. This fraction consists of droplets in the range 400um down to 20um and is known as the spume droplet fraction . Such large particles settle within a matter of seconds rather than minutes, and thus are not transported a significant distance inland.

If we exclude the known corrosive crags from the dataset, then we find that sulphate falls away in a somewhat analogous manner to chloride.

However, if we include just the known corrosive crags in the data set, then the increased prevalence of sulphate is striking, but once again, in a similar manner to chloride, it falls away at distances greater than 1km from the coast.

The way in which the sulphate profile differs from that of chloride is noteworthy. It strongly suggests that the source of sulphur arises from a phase other than that supplying the chloride.

The elemental sulphur hypothesis:

My working hypothesis is that the sulphur source is elemental sulphur of volcanic origin that is transported by the surface biofilm of the ocean. It is deposited with the spume droplet fraction onto the clifftops, where it is converted to sulphuric acid by sulphur oxidising bacteria, an action which contributes exogenous sulphate to the locally circulating ground water.

This biofilm is easily understood as water-insoluble, organic-matter that, being less dense than water, rises to form a thin film on the ocean surface. This interface between the aqueous and organic components on the sea surface provides a rich environment in which micro-organisms can flourish – hence the term biofilm. It is not easily observed unless it is aerated by the action of breaking waves when it forms patches of froth.

A search of the scientific literature fails to come up with any studies regarding the transport of elemental sulphur by the oceans, and certainly not specifically by the surface biofilm, so we really are entering the realms of speculation here. However, there can be little doubt that elemental sulphur, normally totally water-insoluble, can be dispersed by the action of a surfactant , and I can’t think of any argument that might be raised from a fundamental physical chemistry viewpoint to assert that the biofilm couldn’t carry significant quantities of elemental sulphur.

Assuming that elemental sulphur is indeed carried by the surface biofilm, and thus will be swept inshore with the spume droplet fraction, the possibility arises of sulphur being deposited in locations favourable to sulphur oxidising bacteria (SOB) that might be present along ledges and cliff tops. Under the action of these organisms we can anticipate the formation of sulphuric acid, which will migrate with groundwater down through cracks and porosities within the rock. The fact that maximum sulphate levels occur at locations somewhat setback from the ocean is perhaps an indicator that SOB cannot survive directly adjacent to the wave zone but require the more supportive environment of the immediate cliff top.

However, arising out of basic chemistry, we find a substantial objection to this hypothesis. Surely, we’d anticipate that the acid hydrolysis of the dominant alkaline feldspar of the granite would result in an acidic cliff environment marked by elevated levels of potassium and aluminium? However, if such an attack by sulphuric acid does occurs, apart from the presence of residual sulphate, then there is very little evidence for it. In all the samples we have looked at, even where sulphate is elevated, the surface electrolytes are dominated by the pH neutral, carbon dioxide / calcium bicarbonate system.

One plausible explanation for this finding is the fact that, as we will soon discuss, the availability of sulphur is linked to sporadic volcanic activity. It is possible that many years might well elapse between episodes of sulphur deposition, and thus, calcium sulphate being largely insoluble, is the only sulphate likely to be sequestered within the porosities of the rock once the influx of marine elemental sulphur ceases, and the acid environment gives way to a more neutral calcium bicarbonate system.

For now, I should note that, as a point in contrast, the sulphate sea cliffs of Long Dong, Taiwan, present differently. In this case we find acidic ground water depositing pure sodium aluminium sulphate as would be expected from the attack of sulphuric acid on the feldspar components of the sandstone. I’ll write a report on this fascinating corrosive crag at some point in the future.

Marine, elemental sulphur:

Sulphur is a highly conserved element within the biosphere, and the sporadic appearance of sulphur in large quantities is hard to explain except via the agency of volcanic activity.

However, volcanism is restricted either to the boundaries of the tectonic plates, or failing that, to a few well-defined hot spots, as is illustrated in the diagram below.

Simplified plate tectonics for the eastern Atlantic Ocean and the Mediterranean Sea. The plate boundaries are shown as bold red lines. Active volcanoes are marked with the red triangles, whilst loci of seismic activity are marked with orange circles. Source.

Certainly, there are active volcanos within the Mediterranean to the east, as well as within the Atlantic at the Azores and the Canaries to the west, but all of these are well over one thousand kilometres distant and are surely too remote to exert any influence.

Or are they?

Looking for a sulphur source in the Mediterranean:

Let’s take a closer look at the Mediterranean Sea. The illustration below shows that it is comprised of four deep basins. And by deep, I mean seriously deep, reaching beyond 3,000m in many places. The western Algerian and Tyrrhenian basins are interconnected at depth but are substantially isolated from the Atlantic to the west by the Straights of Gibraltar, and from the Ionian and Levantine basins to the east by the Straights of Sicily.

The Mediterranean Sea is comprised of four, deep basins.

The illustration below illustrates how the Mediterranean, although comprised of four basins, can be considered as being just the two distinct bodies of water, one western and the other eastern. Note the depth axis is not linear, and the deeps are far deeper than they appear at first glance. We will move on to discuss the circulation of the various masses of water illustrated in this diagram once we have dealt with the geology.

Of the four basins we see that the Straits of Gibraltar and the Straits of Sicily present shallows that effectively divide the Mediterranean into two substantially isolated bodies of water. Source

Both the Eastern and Western Mediterranean are the result of the same fundamental geological mechanism, i.e., the subduction of the African Plate beneath the Eurasian Plate, but, in spite of this similarity, there are marked differences between the two. Yes, active volcanism is present in both the eastern and western basins, but there are differences which could well bear upon the availability of elemental sulphur.

I am going to have to assume the reader has some knowledge of plate tectonics or we’ll never get done. Hopefully, the diagram below explains the features I wish to refer to. So, let’s think of the righthand side of the diagram as Africa, while on the left is Eurasia. The African Plate is moving beneath the Eurasian Plate to create the four features labelled in bold text across the top, Back-arc basin, Magmatic arc, Forearc and Trench. The Magmatic arc is more commonly referred to as an Island arc, or Volcanic arc. And it should be understood that the Back-arc basin does not necessarily have to develop, but if it does, it can involve extreme spreading of the original crust. The thinning that occurs can be sufficient for new oceanic crust to be upthrust in a manner analogous to that which occurs at the mid-oceanic ridges.

A diagrammatic representation of a subduction zone. In particular, note the four features of Trench, Forearc, Magmatic Arc and Back-arc Basin. I will be using this broad concept to refer to what we see in the Mediterranean basins. Source.

When these broad features typical of the subduction of one continental plate by another are mapped onto the Mediterranean Sea and its surrounds, we see a striking difference between the Eastern and Western basins. See the figure below. Both the Algerian and Tyrrhenian basins of the west can be seen to be back-arc basins whose depth results from extensional crustal thinning processes that have proceeded to such a point that the seabed is comprised of newly formed crust exuded from oceanic spreading ridges (coloured light blue in the diagram below). On the other hand, the Ionian and Levantine basis in the east owe their depth to trenches aligned to the edge of the subducting African plate. The seabed is either old oceanic crust (coloured dark blue in the diagram below), or folded beds of the thrust belt accompanied by accretionary material.

Plate tectonics of the Mediterranean. The light blue regions show the Algerian and Tyrrhenian basins to be back-arc zones where extensional thinning of the crust has proceeded to such an extent that new crust is being introduced at the ridges marked as darker blue jagged lines. Basins in the east, on the other hand, owe their depth to trenches aligned to the edge of the subducting African plate. The seabed is either old oceanic crust (dark blue), or folded beds of the thrust belt (brown) accompanied by accretionary material. Source.

The Algerian Basin, although the result of extensive crustal thinning, has no volcanic arc associated with it. The Tyrrhenian Basin, on the other hand, contains an actively spreading back-arc which sits behind the volcanic arc of the Aeolian Island chain, as illustrated below.

The Tyrrhenian Sea owes its depth to the back-arc extensional basin that is formed behind the Aeolian volcanic arc. The region of newly formed oceanic crust is illustrated. Reference:

Not surprisingly, submarine, hydrothermal activity is recorded at locations across the Tyrrhenian Sea, such as the gas vents off Panarea illustrated below.

Submarine gas vents off the Aeolian island of Panarea. source

And, looking at the active volcanos of the Aeolian arc, we don’t need to look past the crater of Vulcano itself to see abundant deposits of elemental sulphur.

Thus, it is tempting to jump to the conclusion – “Oh look! There is the sulphur source we seek right there!” – but this smoking volcano isn’t necessarily the “smoking gun’ we are seeking in this particular detective story. Ideally, we should be seeing something obvious like that at Kueishantao Island, off the coast of Taiwan. This prolific sulphur emitter, illustrated below, is undoubtedly the cause of the corrosive nature of the sea cliffs of Long Dong.

Now this is a sulphur fumarole! Here we see elemental sulphur being discharged by a submarine vent at Kueishantao Island, Taiwan. There is nothing equivalent to this in the Tyrrhenian Sea. Source

Can we find evidence for such a discharge of elemental sulphur into the Tyrrhenian?

In 2007, John Lupton and coauthors carried out an extensive survey of the seamounts associated with the Aeolian Arc as well as those situated within the back-arc basin. The coverage of that study is illustrated below.

A map showing the extensive coverage of the survey of the South Tyrrhenian Sea by the study of Lupton et al . Whilst evidence for hydrothermal activity was abundant, they found very little to indicate plumes of particulate matter.

The authors used mass spectrometry to calculate the helium isotopic ratios encountered at various depths. As they explain –

They also employed nephelometry to detect plumes of particulate matter such as elemental sulphur or metallic sulphides.

The isotope measurements gave unambiguous indication of hydrothermal contributions across many of the seamounts, but only in one case was a substantial particulate plume identified. Its composition was not identified. So, no “smoking gun”, not really.

An intriguing discovery was that the helium isotope ratio was found to be significantly elevated at deep locations beyond the seamounts raising the possibility of a further deep hydrothermal source.

In summary:

  • The plate tectonics of the South Tyrrhenian Sea favours the subsea introduction of magmatic components such as elemental sulphur.
  • There is abundant evidence of deep hydrothermal connectivity to be found in the helium isotope data.
  • Elemental sulphur is abundant in the output of aerial volcanos of the Aeolian arc.
  • There is no significant evidence of undersea vents releasing sulphur.

Potential transport mechanisms within the Mediterranean:

The Mediterranean basin is characterised by losing more water via evaporation than it gains through estuarine inflow, and thus the waters are markedly more saline than that of the Atlantic Ocean to which it is connected. The less saline waters of the Atlantic enter the Mediterranean as a surface flow through the Straits of Gibraltar, whilst denser more saline water exits the Straights along the bed of the channel.

The diagram below adds a bit more detail to this general picture.

In this diagram of the Mediterranean, we see that the flow to and from the Atlantic is mainly confined to the relatively shallow waters. This overall pattern of less dense water in (blue), over the top of more dense water out (red), occurs without any substantial involvement of the deep water of the two basins. There are deeper mixing mechanisms, but they are on a much smaller scale than the main flow. From .

The inflow from the Atlantic Ocean follows the surface becoming more saline, and warmer, as it approaches the Levantine basin. This body of water is known as the Modified Atlantic Water (MAW) and is shown at a couple of points of its extent in blue.

The outflow occurs below the MAW at intermediate depth and is known as the Levantine Intermediate Water (LIW). It is shown in red in the diagram.

Thermohaline convection between these upper layers and the very deep water occurs in the Eastern Mediterranean, and LIW contains contributions from the so called Eastern Mediterranean Deep Water (EMDW). However, the Western Mediterranean Deep Water (WMDW) doesn’t seem to mix to the same extent. Thus, we have a somewhat isolated body of water that originates in the Gulf of Lion during the cold winter months filling the bottom of the Algerian Basin, before moving down into the Tyrrhenian Basin.

Millot, , suggests that the residence time the WMDW is of the order of a century. The same author, , in a more recent review paper, suggests that the south-eastern Tyrrhenian basin is the point at which the slow turnover of the WMDW is modulated. On those occasions when dense water is supplied from the Ionian Basin, it cascades down the slope from the Straits of Sicily to drive cyclonic currents that sweep through the Western basins to eventually exit at the bottom of the Gibraltar Straits, as illustrated by the red arrows in the diagram below.

Here we see the cyclonically urged path (i.e. ever trying to turn to the right) of the Western Mediterranean Deep Water (WMDW). It enters from the Ionian Basin and exits along the base of the Gibraltar channel. From . Note the potential of this current to entrain sulphur from the Aeolian Arc and transport it to every known sulphate crag of the Mediterranean.

In keeping with the long residence time, we find that the WMDW contribution to the total outflow at the Staits of Gibraltar is but one-third, the rest being supplied by the LIW.

This current sweeping the Western Mediterranean connects the potential sulphur source of Aeolian volcanism to every known sulphate crag.

Potential transport mechanisms beyond the Mediterranean:

The WMDW exits at the bottom of Straits of Gibraltar to become the Mediterranean Outflow Water (MOW). It is a narrowly constrained current that, being denser than the surrounding Atlantic Ocean, would begin to move down the continental slope but for the fact that Coriolis forces drive it clockwise until it tracks the isobaths north-eastward, then northward, along the Portuguese coast.

This behaviour of the MOW has been studied by Price et al, , using spot measurements derived from expendable current profilers. They recorded that the current reached a peak velocity of 1.3m/s and rotated 90º clockwise in 7 hours.

Point vectors for the MOW as it exits the Straights of Gibraltar. The current begins to track the isobaths and thus hug the coast of Portugal. The physics behind this phenomenon is that same as that which causes winds to blow along the isobars in atmospheric circulation. Source .

The illustration below taken from Rogerson et al, , shows the influence of the Mediterranean outflow extends the entire length of the east coast of the Iberian Peninsula, before it turns to track the northern coast.

The cyclonic urging of the potentially sulphur carrying current continues well after exit from the Mediterranean. As a result, Mediterranean-derived water tracks the isobaths up and around the Iberian Peninsula. From the iso-salinity lines it can be seen that this water body is not substantially diluted until is moves north of the peninsula. Ref.

The iso-salinity lines provide a measure of the dilution of the MOW. Note that we must assume that its influence does not diminish significantly until the northern coast of the Iberian Peninsula is reached.

And so, we have a potential linkage between the sulphate crags of Cabo da Roca and the sulphur source of the Tyrrhenian back-arc basin that takes in the sulphate crags of the Western Mediterranean along the way.

Concluding Remarks:

This hypothesis provides a neat explanation for the prevalence of sulphate at specific sea cliffs within the Western Mediterranean, as well as along the extent of Portuguese coast.

However, neat or otherwise, a hypothesis needs to be substantiated with solid evidence before it becomes worthy of consideration. How much evidence do we actually have?

Evidence QualityCategory Details
goodsourceelemental sulphur emitted by Aeolian volcanoes
circumstantialtransportelemental sulphur vented into Tyrrhenian Sea
goodtransportMediterranean currents necessary to connect Aeolian Arc to known sulphate crags
speculativetransportdelivery of elemental sulphur via oceanic biofilm
gooddestinationmeasured high sulphate levels and confirmed SRB attack at multiple crags on both the Portuguese Coast and Ibiza
circumstantialdestinationphotographic evidence of putative SRB attack at multiple crags at Sardinia, Mallorca and Penyal d’Ifac. Needs sampling and confirmation.
How solid is the evidence for this ocean-borne sulphur hypothesis? The weakest link is the delivery of elemental sulphur via the biofilm of the ocean surface. This is nothing more than speculation based on well understood physical properties.

I think the final mark for this thesis should be “work in progress”, but no better. To this end I’ll add that I have made several unsuccessful attempts to find elemental sulphur either on the rock surface or associated with the biofilm. This seems to me to be the most speculative link in the chain. It is also the hardest to verify because it is very likely an infrequent and sporadic event, and lack of evidence cannot be considered evidence of lack.

Rogerson, M., Rohling, E. J., Weaver, P. P. E. & Murray, J. W. Glacial to interglacial changes in the settling depth of the Mediterranean Outflow plume. Paleoceanography 20, (2005).
Millot, C. Circulation in the western Mediterranean-sea. Oceanologica Acta 10.2, 143–149 (1987).
Millot, C. Circulation in the Western Mediterranean Sea. Journal of Marine Systems 20, 423–442 (1999).
Scoon, R. N. Mediterranean Basins and Italian Island Volcanoes. in The Geotraveller: Geology of Famous Geosites and Areas of Historical Interest (ed. Scoon, R. N.) 139–167 (Springer International Publishing, Cham, 2021). doi:10.1007/978-3-030-54693-9_8.
Lupton, J. et al. Active hydrothermal discharge on the submarine Aeolian Arc. Journal of Geophysical Research: Solid Earth 116, (2011).
Steudel, R. & Holdt, G. Solubilization of Elemental Sulfur in Water by Cationic and Anionic Surfactants. Angewandte Chemie International Edition in English 27, 1358–1359 (1988).
Veron, F., Hopkins, C., Harrison, E. L. & Mueller, J. A. Sea spray spume droplet production in high wind speeds. Geophysical Research Letters 39, (2012).
Robinson, A. R., Leslie, W. G., Theocharis, A. & Lascaratos, A. Mediterranean Sea Circulation. in Encyclopedia of Ocean Sciences 1689–1705 (Elsevier, 2001). doi:10.1006/rwos.2001.0376.
Price, J. F. et al. Mediterranean Outflow Mixing and Dynamics. Science 259, 1277–1282 (1993).

7 replies on “Corrosion at Cabo da Roca -6”

This is really interesting! Thanks for this detailed report.

Have you investigated any issues or gotten any samples from Brazil? The UIAA map lists both Rio de Janeiro and Itatim as places where SCC issues may have occurred ( However, I’m not aware of any submarine volcanoes near Brazil’s coast, since it is in the middle of a tectonic plate. Could there be another source of Sulphur?

I saw you investigated 304 bolts from Barra da Lagoa and found them to be ok, but I couldn’t find any mentions to other places in Brazil in your research.

I know that the UIAA corrosion map shows a few locations in Brazil. I really need to see at least photographs and preferably be sent samples before I’ll draw an conclusions.

As you point out there are reasons drawn from plate tectonics to believe that the entire coast of Brazil will be sulphate-free. However, that doesn’t mean there might not be exceptions to this rule. From my point of view, any exception to the sulphate-rule would be fun to investigate, so I am always happy to look at samples if people send them to me.

Sounds good, I’ll contact local climbing organizations and see if I can get my hands in any samples to send to you.

Hi Dave!
This work you’ve done is absolutely incredible. You are a model of how science should be done. Thank. You.
How do I donate to your work, keeping us safer and more educated?
Can we get a Tl;dr?
Here’s my attempt at it; please remove my comment if anything is inaccurate (and post your own!)
The best theory we have so far:
There’s likely an aerobic sulfur-oxidizing bacteria at the top of the crag that turns salt spray into battery acid. But only if the crag is close, but not too close to the ocean, and only if there’s a lot of free sulfur from a volcano nearby.
This battery acid isn’t directly a problem to the stainless.
A second “iron-oxidizing” bacteria forms a biofilm seal around the back of the bolt hanger. This seal allows the acid in and iron ions out, but now oxygen is no longer present.
A third “sulfate-reducing” bacteria takes this battery acid and electrically connects itself to the stainless steel, literally burns it, spitting out a very high pH (base) Fe, and bores little tunnels into only martensite phase of the base metal and shoves both itself, a bunch of sulfur (iron sulfides), and disperses a bunch of hydrogen into atomic dislocations throughout the metal. The hydrogen acts as atomic ball bearings between 3-dimensional slip planes so that when the steel is stressed, it just comes apart.
But. Only if the stainless is relatively cheap, low nickel and low carbon, and was cold worked on a cold day and never post-annealed.

For climbers: if you see a black ring of death behind a stainless bolt hanger at a crag near the ocean, down climb, don’t clip.
If you’re developing a new crag near the ocean, take some tests and send it to Dave before spending money on 304.

Hi Jason, thank you for the kind words.

Thankyou also for the offer of the donation. I’ve been mulling over whether it is worth setting up a system for donations, but running this site is not so expensive if I factor in the amount of fun I’m having. Now that I’m quite well set up, the biggest cost is the carriage for samples. I pay for shipping the sample kit out, and I ask folks to pay for its return.

Yeah, I was aware that TL;DR was an issue, so thank you for your contribution. I reckon that is a pretty good effort, and is reasonably accurate given the complexity of the issue. You have inspired me to do some sort of headline post, and if I may I’ll use some of your words.

Leave a Reply

Your email address will not be published. Required fields are marked *