Dust in Sydney (TG etc)

A paper we should take note of:
Aryal, R., Kandel, D., Acharya, D., Chong, M.N., Beecham., S.  2012.
Unusual Sydney dust storm and its mineralogical and organic characteristics.
Environmental Chemistry 9, 537-546

The dust storm was in 2009 and affected Sydney and Brisbane. Aryal et al (2012) did a thorough study on the dust material, and revealed four particle mode sizes: 0.6, 4.5, 9.3, 20 micrometrres.

The investigators made a neat use of thermogravimetric analysis- we do not see enough TG usage and this application is very welcome. TG revealed an organic content 10.6%. They only report a TG curve, it would have been very useful if they had produced a DTG curve. Their TG curve appears to contain various interesting events.

Our TG reproduction is much better than the original; in the original paper the figures are very small and relatively indistinct. Fig.7 is full of suggestion- a DTG picture could have been remarkably interesting.
The authors concluded that the particles contained Si, Al and Fe in oxide form in which the Al/Si ratio was 0.39. The high organic content and the Al/Si ratio indicated that the particles orginated from agricultural land as well as desert.

Thermogravimetric Analysis of Problem Soils (including Loess)

Thermogravimetric Analysis (TG) is an analytical technique in which the change in weight of a sample is measured as the sample is heated. It is widely used but it is not a popular technique; it is neglected in soil engineering and in engineering geology and (in fact) in all of the earth sciences.

Figure 1 is by George Xidakis (from Smalley & Xidakis 1979) and shows a result for the Modbury Clay (from South Australia; the Modbury High School suffered considerable damage from shrink-swell effects). This sample from CSIRO was analysed on a Leeds University TR02 thermobalance and the result is presented as derivative curves. Fig.1 shows the change in rate of weight loss as samples are heated from room temperature to 800 degrees.

Smalley, I.J., Xidakis, G.S.  1979.  Thermogravimetry of an expansive clay from Adelaide: approximate mineralogical analysis using standard montmorillonites  Clay Science 5, 189-193.


The great advantages possessed by the Stanton-Redcroft TR02 thermobalance were the ability to take quite a large sample, say 1-2 g of soil; a slow rate of heating; results presented in a way to facilitate careful graphical analysis. But the machine did not match the times- by the time the potential of TG techniques in soil engineering had been realised the age of the large sample thermobalance was over. The market required small sample machines for routine analysis which could provide quick heating and cooling. The potential of the large sample thermobalance was never realised- the time might be right for a revival of ground-targeted thermogravimetry.

The pioneering paper was probably Coleman et al (1964) on red soils from Kenya, although the TG applications were not emphasized. The dehydroxylation reaction is very marked in kaolinite (a 1:1 clay mineral) so the kaolinite development in the red soils and laterites of Africa and South America made TG a useful tool. But the first, and most impressive application of TG to a problem soil was in the investigation of the very sensitive post glacial clays of extreme sensitivity (the so-called quickclays).

Coleman, J.D., Farrar, D.M., Marsh, A.D.   1964.  The moisture characteristics, composition and structural analysis of a red clay soil from Nyeri, Kenya.  Geotechnique 14, 262-276.

Fig.3 is the derivative plot for the St.Jean Vianney clay. SJV was the site of a significant quick clay landslide (several killed, much damage) and this resulted in much investigation of the ground material. The SJV DTG plot is quite complex. A1- adsorbed water: a small peak- the clay minerals are of the inactive nature. The equivalent peak for the Modbury clay is large, the Modbury clay is dominated by smectite type clay. B2 organic material; C1 dehydroxylation reaction- but muted, not a large amount of an inactive 2:1 clay mineral; D carbonate breaks down to CaO + CO2. This particular result showed that there was a small amount of clay mineral material in the SJV clay, the ground material was not really 'clay' - the SJV clay was dominated by clay-size primary minerals; the clay minerals played a relatively small role in determining the properties. The sensitivity was not due to clay mineral content.

Smalley, I.J., Moon, C.F., Bentley, S.P.  1975.  The St.Jean Vianney quickclay. Canadian Mineralogist 13, 364-369.

This is a particularly satisfying picture from Smalley et al 1975. This is the SJV quick clay (note the alternative terminologies- there was a move to call the quick clays quickclays to show that they were distinctive materials- to detach from the clay label)  2 g samples showing -at point C - the dehydroxylation reaction of the inactive type clay- probably illite. The carbonate peak is well defined- a beautiful piece of analytical architecture.

TG/DTG techniques were appreciated in Hungary; a whole series of standards was produced and published as loose-leaf cards. (see G.Liptay 1971  Atlas of Thermoanalytical Curves, Akademiai Kiado Budapest). Card 123 was Potassium Hydrogen Phthalate: recommended for use as a thermal standard.