The Geochemistry of the Red Clays of Holderness and their origin.

Extra data is available on the CD-ROM of Humberside Geologist number 14.

Abstract.

This research was carried out as a forensic science undergraduate project at the University of Hull, Department of Chemistry. Samples of red clay (with no gravel or pebble content) and red boulder clay from the coast of Holderness were compared with clays of known geological age and origin. The techniques used included ICP analysis, SEM analysis, acid carbonate testing, microscopy and particle sizing. The red boulder clays showed similarities with a weathered boulder clay. The red clays showed similarities with unweathered boulder clay. They showed no affinity to the Red Chalk or Mercia Mudstone.

Introduction

This project was trying to ascertain if it was possible to prove conclusively whether one or more rocks came from the same source by using currently existing analytical methods and instruments. To answer this question several different samples of red clays were selected for use in this experiment.

Geological background

The Boulder Clays (or Tills) of Holderness were deposited by the Ice Ages of the late Quaternary by glaciers that had travelled from the Lake District, Scotland, Northern England and Scandinavia. We have evidence for this from the erratic rocks and fossils that can now be found in Holderness (Horne & Harrison 1992, Horne 2000, Rockett 1992).

Thin layers of red and grey clay and soft chalk can be seen between the layers of the Tills in the cliffs of Holderness or occasionally as beach exposures (Bisat in Catt and Madgett 1981, Fenton 1977). The chalk in these 'rafts' is younger that the chalk exposed in East Yorkshire, and has probably been picked up by the glaciers as they moved across the bed of the North Sea (Horne, ongoing research). The grey clays seem to be of Late Jurassic age, from evidence collected by Stuart Jones and Chris Brogden. But what is the origin of the bands of red clay and red boulder clay? There are several possibilities :-

• red rocks of Devonian age
• red mudstone of Triassic age from the North Sea
• Red Chalk of Early Cretaceous age from the North Sea
• red clays of Tertiary age (not seen on-shore)
• red boulder clay from earlier in the Quaternary
• an older Boulder Clay that has weathered to a red colour and then been scraped up and re-deposited by a glacier
• weathering of the lower Boulder Clay to a red colour before being covered by another layer
• red clays deposited by fresh water in the North Sea area when sea level was low
• red clays deposited in shallow marine conditions in the North Sea area when sea level was low

Samples of red clay from a beach exposure at Aldbrough, East Yorkshire [sample numbers ALD2005-3, ALD2005-5, ALD2005-VR] and Red Boulder Clay from the "Red Band" in the boulder clay cliffs at Aldbrough [ALD2005-2] , Hornsea [HORNRBC] and Mappleton [MAP2005] were collected by Anne Horne, Mike Horne and Gordon Ostler in 2005. These were compared with other samples of red clays and boulder clay.
 

Sample number	Geographical Location	Geological Age	Description
ALD-2005-2	Aldbrough (NGR TA260392 )	Pleistocene	Red boulder clay
ALD-2005-3	Aldbrough(NGR TA260392 )	Pleistocene	Red clay
ALD-2005-5	Aldbrough(NGR TA260392 )	Pleistocene	Red laminated clay
ALD-2005-VR	Aldbrough(NGR TA260392 )	Pleistocene	Very red clay
BOOKBARN	Book Barn, Hallowtrow, Somerset. (NGR ST6356 )	Triassic	Red clay - Mercia Mudstone Group
HOLMEOSM	Church Hill, Holme-On-Spalding-Moor. (NGR SE819388)	Holocene	Red Soil - ?derived from Mercia Mudstone Formation or an early glaciation.
HORNBC	Hornsea (NGR TA2118446941 )	Pleistocene	Boulder Clay (unweathered)
HORNRBC	Hornsea(NGR TA2118446941 )	Pleistocene	Red Boulder Clay
HORNWBC	Hornsea(NGR TA2127746629 )	Quaternary	Weathered Boulder Clay
MAP2005	Mappleton(NGR TA230434 )	Pleistocene	Red boulder clay
SE452766	Sugar Loaf Bay, Portishead, Somerset. (NGR SE452766)	Devonian	Red clay from the Old Red Sandstone.
SPEETON	Speeton (NGR TA153755 )	Albian, Cretaceous	Red marl from Red Chalk Formation.

 

Most of the samples were collected from coastal locations, except SE452766, HOLMEOSM and BOOKBARN.

Techniques used -

[safety note - the specialist chemical techniques were carried out in a science laboratory and for safety reasons should not be repeated at home!]

ICP

Inductively Coupled Plasma- Optical Emission Spectrometry (ICP-OES) is a solution based technique for elemental analysis. An ICP is a very high temperature (7000-8000 K) excitation source that efficiently desolvates, vaporises, excites and ionises atoms.

The most essential step in analysis is the sample preparation; as unless appropriate sampling schemes and suitable procedures are used to prepare samples any subsequent analysis will be wasted. The sample needs to be ground up to give a small particle size and then dissolved in acid. A weak acid is generally referred to as a leach and while it may result in a quantitative extraction of the elements of interest, most of the sample will remain undissolved. Strong acid attacks are generally referred to as digestions and these are more powerful than leaches. For the experiment carried out in this project, aqua regia digestion was used. Aqua regia is a 3:1 mixture of hydrochloric and nitric acids. Nitric acid destroys organic matter and oxidises sulphide material. It reacts with concentrated hydrochloric acid to generate aqua regia according to the equation:-

 3HCl + HNO₃ ==> 2H₂O + NOCl + Cl₂

Aqua regia is an effective solvent for most base metal sulphates, sulphides, oxides and carbonates but only provides a partial digestion for most rock forming elements. This acid digestion is greatly speeded up when combined with microwave digestion. This prepares samples at high temperature in closed vessels and allows smaller quantities of acid to be used. It also ensures that certain inorganic compounds that are quite resistant to acid attack are digested. The digested sample can then finally be diluted and analysed. This analysis could then give detailed information on the chemical composition of the samples.

An aqua regia solution (100 ml) was made by mixing concentrated hydrochloric acid (75 ml) and concentrated nitric acid (25 ml). The solution was left to stand overnight in a fume cupboard to allow the chlorine gas to escape and the bright orange colour to develop. Each sample (0.3 g) was added to the digestion containers and aqua regia solution (5ml) added before the containers were loaded into the microwave digestion machine. This digestion was carried out at 80 p.s.i. for 10 minutes and then the samples left to cool. The samples were removed and added to a 25 ml conical flask ensuring thorough rinsing of the containers and then the flasks made up to the mark with water. These samples were diluted further by taking 1ml of sample and adding it to a 10 ml conical flask and made up to the mark with 2% HNO₃. The samples could then finally be analysed using the ICP-OES instrument.

Scanning Electron Microscopy

SEM analysis involves directing a beam of electrons at the sample. Electrons are emitted from the sample and collected by a detector which converts them into a small electric signal. This signal contains a variety of information about a single point on the sample surface. To form an image of the sample a large number of points over an area need to be analysed, and the final image is built up from the number of electrons emitted from each point. To enable all of this to occur, samples need to be prepared carefully to withstand the vacuum inside the microscope but they also need to be made conductive. This is achieved by sputtering or vacuum evaporation to put a thin carbon or metallic coating onto the sample.

Scanning electron microscopes are often coupled with x-ray analysers. This is because x-rays are produced whenever an electron beam interacts with matter and these can be used very effectively to give information about the chemical compositions of the sample. Energy Dispersive X-ray micro analysers (EDX) are the most commonly used as these can detect almost all elements simultaneously and these elements are identified from their characteristic energies and their concentration can be derived from the count rate. This chemical composition is achieved using bulk analysis rather than spot analysis. This is because bulk analysis gives an area analysis and so the best overview of the sample, whereas spot analysis is best used for analysing a specific grain.

The sample pellets were produced by the method and equipment normally used to make K Br discs for Infra-red analysis. The powdered samples and pellets were then sent for analysis. The analyses represent an average of five spot samples.

Acid Carbonate Testing

A recorded amount (about 25g) of each of the powdered rock samples was taken and mixed with a small amount of de-ionised water and then dilute hydrochloric slowly was added. Once the samples had stopped fizzing more acid was added until the reaction with the samples was complete. The samples were filtered to remove the liquid and the remaining solids transferred to filter paper and dried. After thorough drying the samples could be re-weighed and the carbonate content calculated. The acid reacts with the calcium carbonate in the rock samples and hence the weight of the rocks, once the acid has been filtered off, decreases providing an indication of the amount of carbonate in each of the samples.

Particle Sizing

For this analysis the samples were first boiled in water with a de-flocculating agent to break the rock down into a slurry. This slurry was put through a set of sieves to filter out the minerals and bits of rock in the sample, and this allowed all the silt and clay to be removed. The set of sieves used had the following mesh sizes: 63um, 250um, 500um and 1mm.

• Sample is weighed, from 300 g to 1300 g and placed in metal basin
• The basin was then filled with cold water
• Added to the water was 5 g of sodium hexametaphosphate, this was added to deflocculate the clay and to clean the sample
• The sample was then boiled for 40 minutes, with stirring
• The samples were then washed into a 63 um sieve and washed through until the water ran clear
• If any rock was not broken down in the first boil then re-boiling is needed, a maximum of three re-boils
• The samples are then washed through a stack of sieves from size 1 mm, 500 um, 250 um and 63 um.
• The different fractions are then washed into beakers
• The samples were then filtered by vacuum and the sample was dried in an oven at 60°C

Once the particle sizing had been carried out microscopic analysis of the fractions could be done to see what grains, and possibly minerals were present. This part of the research used the particle-sized samples: a selection of the samples from the 500 um -1mm and 250 um -500 um particle size ranges were looked at under a microscope. The different types of particulates seen were categorised and counted. About 200-300 particulates were counted for each sample; this was to ensure that the results obtained gave a true representation of what each sample contained.

Results and discussion -

ICP analysis -

ICP analysis provided the best results of the project in terms of answering the question of whether the samples could be matched to one another. Although it could not provide a definite answer that two samples were identical, it did show there were two groups of similar compositions, and one group with significantly different compositions to the rest. 

Group A

BOOKBARN, SPEETON, SE452766 and HOLMEOSM.

BOOKBARN and SPEETON are very obviously different to the other samples. BOOKBARN contains a very large amount of iron whilst SPEETON is mostly made up of calcium. SE452766 and HOLMEOSM have elemental compositions that are very similar but are from opposite ends of the country, with SE452766 originating from Somerset and HOLMEOSM from Yorkshire. Therefore these two samples are unlikely to have originated from the same source even though they are elementally similar.

Group B

HORNBC, ALD-2005-5, ALD-2005-3 and ALD-2005-VR.

All of these samples have virtually the same elemental composition. They contain about 5% calcium and have a significant amount of iron in them. HORNBC was essentially used as a standard as it was known to be a boulder clay. Therefore the conclusion can be drawn that the other three samples in the group are most probably also boulder clays and originate from the same area/source.

Group C

HORNWBC, ALD-2005-2, MAP2005and HORNRBC.

These four rocks are compositionally similar; they contain roughly an equal ratio of iron and calcium. HORNWBC is known to be a weathered boulder clay and in comparison to HORNBC (boulder clay) has less calcium but more iron present. This could show that the weathering slightly alters the percentage elemental composition of rocks. Hence, HORNRBC, ALD-2005-2 and MAP2005could originate from the same area/source as HORNWBC or it could just be rocks that have undergone a similar weathering process.

Although these differences cannot be explained in terms of locations, it does provide evidence that samples have similarities in their various compositions. It may also be an indicator that they possibly originate from the same areas before the glacial movements of the Ice Age.

 SEM Analysis -

Morphological information

Scanning Electron Microscopy (SEM) provided images of the powdered samples at a micrometre size, as well as a basic elemental composition of the sample. Pictures of all 12 of the samples were taken using the scanning electron microscope; each sample was photographed at three different magnifications.

The morphological information is seen the SEM pictures. These were taken at three different magnifications for each sample. To obtain the pictures a smooth surface needs to be achieved which is done by grinding the sample into a fine powder. Unfortunately this destroys some of the surface characteristics and so an accurate comparison of the samples cannot be carried out. It is also difficult to carry out a comparison due to the samples all being quite similar and there being no characteristic features such as fossils being present.

Compositional information

Using an X-ray detector, an elemental breakdown on the samples was also possible using the SEM apparatus. The compositional data was obtained by randomly analysing five spots on the sample and then taking an average of the results. The results of this are summarised in the tables below:

Raw Data
 

Sample	Na	Mg	Al	Si	S	Cl	K	Ca	Ti	Fe	O
SPEETON	0.41	1.05	9.19	19.8	0.05	0.59	2.26	18.4	0.43	6.25	41.54
SE452766	0.24	3.26	10.22	25.68	0.15	0.9	4.64	2.71	0.5	6.63	45.07
MAP2005	0.72	1.77	10.72	26.87	0.11	0.22	3.42	2.9	0.64	6.69	45.93
HORNWBC	0.45	1.77	12.71	25.08	0.3	0.1	3.07	4.14	0.54	5.88	45.97
HORNRBC	0.48	1.78	12.68	25.7	0.28	0.86	3.09	2.23	0.62	6.24	46.04
HOLMEOSM	0.31	5.46	8.97	23.65	0.05	0.04	3.92	6.59	0.48	6.27	44.25
BOOKBARN	0.2	1.2	9.02	28.92	0.05	0.01	5.87	0.52	0.55	7.76	45.9
ALD-2005-VR	0.81	1.59	10.22	24.51	0.14	0.5	3.19	7.9	0.52	6.15	44.47
ALD-2005-5	0.78	1.51	9.36	26.06	0.21	0.56	2.93	7.49	0.6	5.38	45.12
ALD-2005-3	0.75	1.53	9.43	25.19	0.18	0.4	2.96	8.72	0.55	5.6	44.69
ALD-2005-2	0.94	1.73	10.51	25.99	0.07	0.53	3.67	3.79	0.55	7.05	45.17
HORNBC	0.63	1.59	10.99	24.11	0.46	0.93	2.8	7.4	0.5	5.86	44.73

Percentages without oxygen and silicon:-
 

Sample	Na	Mg	Al	S	Cl	K	Ca	Ti	Fe
SPEETON	1.06053	2.71599	23.77134	0.12933	1.52613	5.84584	47.67201	1.11226	16.16658
SE452766	0.82051	11.1453	34.94017	0.51282	3.07692	15.86325	9.26496	1.70940	22.66667
MAP2005	2.64803	6.50975	39.42626	0.40456	0.80912	12.57815	10.66569	2.35381	24.60463
HORNWBC	1.55387	6.11188	43.88812	1.03591	0.34530	10.60083	14.29558	1.86464	20.30387
HORNRBC	1.69851	6.29866	44.86907	0.99080	3.04317	10.93418	7.89101	2.19391	22.08068
HOLMEOSM	0.96603	17.0146	27.95263	0.15581	0.12465	12.21564	20.53599	1.49579	19.53880
BOOKBARN	0.79428	4.76569	35.82208	0.19857	0.03971	23.31215	2.06513	2.18427	30.81811
ALD-2005-VR	2.61122	5.12573	32.94649	0.45132	1.61186	10.28369	25.46744	1.67634	19.82592
ALD-2005-5	2.70645	5.23942	32.47745	0.72866	1.94310	10.16655	25.98890	2.08189	18.66759
ALD-2005-3	2.49004	5.07968	31.30810	0.59761	1.32802	9.82736	28.95086	1.82603	18.59230
ALD-2005-2	3.25936	5.99861	36.44244	0.24272	1.83773	12.72538	13.14147	1.90707	24.44521
HORNBC	2.02182	5.10270	35.26958	1.47625	2.98460	8.98588	23.74840	1.60462	18.80616

Notes - All samples were collected from the beach or low cliffs at coastal exposures, except HOLMEOSM and BOOKBARN which are inland, and HORNWBC which was from the top of a cliff section: the chlorine levels probably reflect the influence of sea water.

It can be seen that all of the samples had very similar silicon (approx 25%) and oxygen (approx 45%) content, but once these elements were removed the charts started to show a large correlation to the ICP results. From the SEM compositions the samples can be grouped in the same way. Group A are the samples that are very different to the others and it can be seen again that the SPEETON sample has a very large calcium content, whilst BOOKBARN has very little calcium but large amounts of iron. The weathering effect on samples can also be seen with these results. The calcium content decreases between group B and C, but an increase in the iron content was not seen with the SEM results as was seen from ICP.

 Carbonate testing 

% carbonate = ((original weight - weight after digestion) /original weight) x 100

The acid test showed that all of the samples contained some carbonate but in varying amounts. This was expected due to all of the samples being soils and so most probably containing some quantity of chalk.

SPEETON contained a very large amount of carbonate; this quantity was much larger than that in any of the other samples. By contrast, BOOKBARN contained very little. It contained only 10% of the quantity in the SPEETON sample. These results therefore correlate with those found from the ICP analysis.

Particle sizing

Particle sizing was carried out for seven of the samples and the following results obtained:- 

Another surprise was that HORNRBC showed the smallest amount of particles under 63 um, as this was assumed to be the case for HORNWBC – the weathered sample. An explanation however could be correlated with the colour of the sample. As ICP analysis found, the majority of the samples had a large percentage of iron in their composition, and the oxidation states of iron give red colours i.e. rust. A possible explanation therefore is that this sample had a degree of oxidation to it. This may have been due to water or weathering whilst it was still in formation during the Ice Age, and consequently some of the material may have broken down.

The fractions were placed into three groups according to their amounts within the 250 -500 um range. These were:

a) BOOKBARN, MAP2005, and HORNWBC

b) HORNRBC and HORNBC

c) HORNVRC and SPEETON.

Unfortunately, these fraction percentages appeared to be the only similarities between the samples, as when comparing these groups for the other fraction similarities there are none.

Microscopy

Once the particle sizing had been carried out microscopic analysis of the fractions could be done to see what grains, and possibly minerals were present. The particle size 250 - 500µm was mainly studied and the following results obtained:- 

The results form the optical microscopy are very varied from sample to sample, and again no patterns can be spotted between the samples analysed. This may again be due to the limited number of samples analysed in this matter, as only half of the 12 samples were used, and so the spectrum of samples was not as big as for the other methods.

Another big problem is the classification system employed in this procedure. As none of the people that undertook this technique had much geological knowledge the classification system used relied mainly upon colour, size, shape and lustre. This means that the analysts used some personal discretion. For example what one person may believe is a smoothed black particle, could to another scientist be seen as a rounded dark grey particle. Basically, due to the specialist nature of this technique it could not be performed to its full potential.

The results that were found were placed into three groups based on the percentage of ‘clear quartz’ counted within the samples. This method would have been more beneficial to the project if it had been carried out by a geologist. It was found that although not exact, the two different sizes of HORNBC did have a rough similarity, which was expected considering the sample itself did not change in any way. It was interesting that based on this grouping there was some degree of matching possible; as the results show the BOOKBARN and SPEETON samples are grouped together for the ICP results and appear to have a similar number of clear quartz crystals; and MAP2005and HORNWBC also follow this trend. All four however from this analysis do have similar particle compositions though, hence the findings were considered inconclusive.

The three groups were:

a) HORNBC at 250 and 500 um

b) ALD-2005-VRC 250 um and HORNRBC 500 um

c) BOOKBARN and SPEETON at 500 um.

Conclusions

 The SEM analysis has managed to produce some results extremely similar to those produced by the ICP. As an example of how well the results seem to compliment each other, the results for the amount of calcium found in the samples by ICP and SEM are compared below with the carbonate content:- 

So, the SEM results and ICP results appear to agree on the elemental compositions for most of the elements.

The acid digestion results also seem to be correct when used in conjunction with the ICP and SEM data collected on the amount of calcium in the samples, as can be seen in the table to the left as well. There is clearly a correlation between the percentages in all three of the techniques; they all increase with respect to each other in most cases. It does appear that the SEM-EDX may not be as sensitive as the ICP and the acid digestion calculation is slightly off for some of the samples. This could be due to the acid dissolving more than just the calcium in the sample, and so the calculated values may not be completely accurate.

The samples analysed seem to fall into three groups; in two of the groups the samples are very similar chemically -

A - A diverse group that show no particular similarities to the other groups. The Holme-on-Spalding-Moor sample [HOLMEOSM] is a clay soil from the Vale of York: could the high magnesium level be from an influence from Magnesian Limestone in a glacial deposit, or is the higher potassium and magnesium from agricultural fertilisers? The marl from the Speeton Red Chalk [SPEETON] has a high calcium carbonate content. Previous studies have revealed no microfossils of Albian age (Horne unpublished) in the Red Boulder Clay bands. The Mercia Mudstone [BOOKBARN] showed high iron content and low calcium carbonate.

B - ALD-2005-3, ALD-2005-5, ALD-2005-VR and HORNBC. The three red clays from the beach at Aldbrough have a similar chemical composition, as well as similar grain size distributions. They are similar to the chemistry of the un-weathered boulder clay from Hornsea. Could they have been washed out of an unweathered Boulder Clay? There are also some similarities with the Old Red Sandstone sample from Somerset which are hard to explain.

C - ALD-2005-2, HORNRBC, MAP2005 and HORNWBC. The three red boulder clay samples were collected from sites several kilometres apart and show great similarities with the weathered Boulder Clay. They are from the Red Band between two Boulder Clays and could therefore represent a layer of weathering or a smear of weathered boulder clay pushed along by the glacier.

Overall though the data collected from the three more chemical processes it appears that many of the results seem to support the three groups theory. The physical processes employed turned out to be a little disappointing especially the microscopy, as it was hoped that a comparison between the SEM pictures and what could be seen down an optical light microscope could be compared but unfortunately this was not possible. The particle sizing also gave no relevant information.

Acknowledgements -

Thanks to the project supervisors Tony Walmsley and Mike Horne, Tony Sinclair for his help with the SEM analysis, Bob Knight for his help in the ICP analysis, Marion Brazier and Mark Anderson of the Geography Department for the loan of equipment and the Chemistry Department technicians in the laboratory for their helpfulness throughout this project. The project was funded by the University of Hull Department of Chemistry.

References and bibliography -

Catt J A & Madgett P A 1981. The work of W S Bisat F.R.S. on the Yorkshire coast. pp 119-136 of Neale & Flenley The Quaternary in Britain. Pergamon Press. 267 pp.

Fenton K 1977. Bridlington to Hornsea. Humberside Geologist. 2, 10 (not numbered).

Hill S J (ed.) 1998. Inductively coupled plasma spectrometry and its applications.

Horne M & R Harrison 1992. The East Riding Boulder Committee, reports for the years 1987 to 1991. Humberside Geologist 10,18-22.

Horne M 2000. Report of the East Riding Boulder Committee 1992 to 2000. Humberside Geologist 13, 42-45.

Horne M, Rockett T & Whitham F 2000. The Glacial Geology of Dimlington High Cliff Humberside Geologist 13, 53-56.

Horne M 2005. Hull Geological Society’s standard methods for recording Quaternary sediments (proposals). <http://www.horne28.freeserve.co.uk/flamsop.htm>

Lawes G 1987 Scanning electron microscopy and X-ray microanalysis. John Wiley & Sons.

Perrin M B 1975. An Introduction to the Chemistry of Rocks and Minerals. Edward Arnold.

Pye K & D J Croft (eds.) 2004. Forensic geoscience: principles, techniques and applications Geological Society [London] Special Publication no. 232. 318pp

Reeves R.D.& R.R. Brooks 1978. Trace Element Analysis of Geographic Materials. John Wiley and Sons.

Rockett T 1992. Glaciation and the YorkshireCoast. Humberside Geologist 10, 14.

Smith K A 1983. Soil Analysis: Instrumental Techniques and Related Procedures. Marcel Dekker Inc.

Weaver C E 1989. Clays, Muds, and Shales Elsevier.

Winefordner J D (ed) 1996 Introduction to x-ray powder diffractometry. John Wiley & Sons.

 
(c) Hull Geological Society 1999 + 2007

Humberside Geologist Online