Hagerman, UC18699

Identification of Soluble Salts in a Ceramic Jar from the

Petrie Museum of Egyptian Archaeology

Figure 1. UC18699. Jar from the Petrie Museum of Egyptian Archaeology.

Abstract
How can conservators non-invasively identify salts on ceramics with small sample sizes of spalled
material, which is mixed with pottery? How can the salts be analysed without the use of water
prior to desalination? These questions are especially applicable to ceramic objects from the Petrie
Museum of Egyptian Archaeology. Many of these artefacts have a significant soluble salts
problem. UC18699 from the Petrie Museum is the focus of this study (Figure 1, above). Soluble
and insoluble salts of unknown composition have effloresced on its surface, causing structural
cracks and spalling and delamination throughout the ceramic. These salts could potentially be
from the jar’s use to hold offerings or from its burial in the tomb from which it was excavated. A
variety of analytical methods: wet chemical spot tests, pXRF, SEM-EDS, and XRD were
employed to determine the composition of the salts using only a small sample of spalled material.
To assist with this aim, experiments were conducted with terra cotta pots and mixtures of known
salts under the XRD. This experiment was an attempt to determine the accuracy of the
instrument in determining composition values in a mixed sample. Knowing the composition of
salts in UC18699 informed the desalination process. If only insoluble salts had been present, the
decision would have been made not to desalinate at all.
Table of Contents
I. INTRODUCTION 4
1. UC18699 AND SOLUBLE SALTS 4
1.1 UC18699 4
1.2 THE PROBLEM: SOLUBLE SALT IDENTIFICATION 5
1.3 LITERATURE REVIEW 6
1.3.1 DESALINATION AND CERAMICS 6
1.3.2 COMMONLY USED INSTRUMENTATION 7
1.4 OBJECTIVES 9
1.5 RELEVANCE AND IMPORTANCE 10

II. MATERIALS AND METHODS 11

ANALYSES PERFORMED 11

III. RESULTS 14
1. VISUAL ANALYSIS 14
2. SPOT TESTS 15
3. PORTABLE X-RAY FLUORESCENCE 16
4. X-RAY DIFFRACTION 17
5. SEM-EDS 20

IV. DISCUSSION 23

V. CONCLUSIONS 25

VI. ACKNOWLDGEMENTS 27

VII. REFERENCES 27

VIII. APPENDIX 29

2
List of Figures and Tables
Figures
Figure 1. UC18699. Jar from the Petrie Museum of Egyptian Archaeology.
Figure 2. The case where UC18698 and UC18699 are displayed in the Petrie Museum of
Egyptian Archaeology. UC18698 is visible on the middle shelf, far left.
Figure 3. UC18698 (left) and UC18699 (right) side by side. Shows the difference in soluble
salt efflorescence and condition between the two "twin" jars from Harageh. (Petrie
Museum, 2014).
Figure 4. Detail of cracking (top) and salt efflorescence, friable surface (bottom).
Figure 5. Rates of solubility among commonly found soluble salts. Adapted from Koob &
Ng, 2000.
Figure 6. UV Dino-Lite shows small fluorescent dots in the surface of UC18699
Figure 7. Picture shows where pXRF tests were performed on UC18699. Location 1: White
surface deposits. Location 2: Area of the surface without visible salt efflorescence.
Location 3: Gray surface deposits.
Figure 8. XRD spectra of small mixed sample, pottery and soluble salts. Shows likely
presence of sodium chloride, NaCl (blue) and quartz silica, SiO2 (red).
Figure 9. XRD spectra of known sodium chloride and terra cotta mixed sample. Salt and
pottery were mixed in a 1:1 ratio by weight.
Figure 10. XRD spectra of known sodium chloride, calcium carbonate, sodium sulphate
(anhydrous), and terra cotta sample. Salts and pottery were mixed in a 1:1:1:1 ratio by
weight.
Figure 11. XRD spectra of known sodium chloride, calcium carbonate, sodium sulphate
(anhydrous), and terra cotta spectra (blue) overlaid with UC18699 spectra (red). Salts and
pottery were mixed in a 1:1:1:1 ratio by weight.
Figure 12. SEM images of soluble salts (top, middle) and close-up view of pottery (bottom).
Figure 13. Photos of the desalination process. Top: UC18699 is fully submerged in deionized
water. Bubbles forming on the surface show the movement of air and salts out of the
ceramic body. Bottom: the desalination set-up.
Figure 14. UC18699 after desalination.

Tables
Table 1. Specifications of instrumentation to be utilized in this study.
Table 2. Table adapted from Johnson, 1998, 2. Spot tests from Odegaard et al., 2001.
Table 3. Experimental mixtures to be run through the XRD.
Table 4. Images of UC18699 obtained using the Dino-Lite (Pro HR AM-7013MT).
Table 5. Results of spot tests performed on spalled material.
Table 6. pXRF: Showing the major and minor elements, Location 1.
Table 7. pXRF: Showing the major and minor elements, Location 1.
Table 8. pXRF: Showing the major and minor elements, Location 3.
Table 9. Results of the elemental analysis under the EDS shows the major and minor elements
in each area of analysis. Full spectra in Appendix _, Figure, .
Table 10. Results of the elemental analysis under the EDS shows the major and minor
elements in each area of analysis. Full spectra in Appendix _, Figure, .
Table 11. Shows elements present in UC18699 as found by completed analyses (X). Presence
of carbonates is marked (I) for inconclusive.

3
I. Introduction
1. UC18699 and Soluble Salts
1.1 UC18699
The surface of UC18699 (Figure 1), a jar in the collections of the Petrie Museum
of Egyptian Archaeology, is covered in white and grey accretions. Originally from
Harageh, a tomb group site in Northeastern Egypt (Egypt D-K/Egypt), the pink ware jar
is dated to the Middle
Kingdom, 12th Dynasty
(Engelbach, 1923). Under the
supervision of Sir Flinders
Petrie, archaeologists Reginald
Engelbach and Battiscombe
Figure 2. The case where UC18698 and UC18699 are displayed in Gunn excavated the jar and
the Petrie Museum of Egyptian Archaeology. UC18698 is visible
on the middle shelf, far left. many others like it in the
winter of 1913-1914 (University College London, 2001). A drawing of the jar can be
found in Engelbach, 1923, pp. 10; Plate XXXIX, Figures 67O and 67O2 (Appendix,
Figure 1).
The jar is roughly cylindrical, (21 x 14 centimetres) with thick walls (0.9 cm).
Whorl patterns on the surface and base of the object suggest the jar may have been made
on a wheel. The most striking
features of the jar, however, are the
salt accretions. They cover much of
its surface as well as two long cracks
stemming from the rim. The
efflorescence is concentrated on the
side of the jar with the label
“U.C.18699” and the bottom of the
vessel.
The object has a “twin” in
the Petrie Museum’s collections,
UC18698 (Figure 2; Figure 3). Figure 3. UC18698 (left) and UC18699 (right) side by side.
Shows the difference in soluble salt efflorescence and
According to a blue inscription condition between the two "twin" jars from Harageh.
(illegible on 18699) written by (Petrie Museum, 2014).
Petrie, both objects held offerings to the Pharaoh Senusret II (Petrie Museum 2014).

4
Harageh, the publication of the excavation, notes that both jars may have held natron
(Engelbach, 1923). Natron is composed of a mixture of mainly Na2CO3•10H2O and
NaHCO3 with NaCl and NaSO4 in small amounts (Abdel-Maksoud & El-Amin, 2011).
Unlike UC18699, UC18698 does not display any visible salt efflorescence (Figure 3).
According to Petrie Museum
conservator, Susi Pancaldo, UC18698
was desalinated after a new roof was
installed in the 1990s (Pancaldo, 2014).
It is unknown why UC18699 was not
washed at the same time. Perhaps
soluble salts had not yet effloresced on
its surface.

1.2 The Problem: Soluble

Salt Identification
Soluble salts expand and
contract in the ceramic body of vessels
in response to fluctuations in relative
humidity (Buys & Oakley, 1993; Ellis, et
al., 2007; Johnson, 1998). Over the years
Figure 4. Detail of cracking (top) and salt efflorescence, the deliquescence of soluble salts in the
friable surface (bottom).
body of the ceramic has caused the
surface of the vessel to spall and crack. The surface is friable and spalls off when handled
with gloves. Ceramic and salt dust has collected in the packaging material. Crystals can be
seen inside cracks around the rim and base. The cracks in the rim move slightly when
light pressure is applied. The areas of loss on the base have rendered the bottom of the
vessel thin and weak.
To prevent further loss, the jar was desalinated in January 2015. Many studies
utilise salt crystals, crystallised from wash water to determine the composition of soluble
salts. This method however is very labour-intensive. This would not have been feasible
on a hot plate in the conservation lab. Instead, a small amount of material has been
collected from the packaging material. This provided all the sample size for destructive
forms of analysis.
This study hopes to identify the salts present prior to desalination in a non-

5
destructive manner to preserve the already friable surface of the jar. The small sample
(approximately 0.386g) will be all the material available for analytical tests. Different salts
have different solubilities and rates of solubility based on their chemical composition and
polarity (Buys & Oakley, 1993; Koob & Ng, 2000; Figure 5). Insoluble salts include
phosphates and carbonates, which do not dissolve readily in water at standard
temperature and pressure (Johnson, 1998; Table 2).

Slower Faster

Sulphates Chlorides Nitrates

Figure 5. Rates of solubility among commonly found soluble salts. Adapted from Koob & Ng, 2000.

Identifying the salts present in the mixture influenced the decision to desalinate
the vessel. If only insoluble salts had been present, soaking the vessel in water for hours
or days would have exposed the jar to a wholly unnecessary risk of damage.

1.3 Literature Review

Many authors write about the identification of soluble salts after desalination
(Ellis, et al., 2007; Koob & Ng, 2000; Odegaard, et al., 2007; Paterakis, 1987; Riss, 1993).
These tests are carried out directly on wash water from desalination or using crystals
obtained by evaporating the wash water instead of on crystals sampled from the object.
Many studies concerning soluble salts concentrate on their effect on ceramics (Gazulla, et
al., 2010; Jin, et al., 2014; Pessoa, et al., 1996; Unruh, 2001; White, et al., 2010). Other
studies focus on the identification of soluble salts on related materials such as lime
mortar and brick (Gonçalves, et al., 2006), limestone sculptures (Bradley & Middleton,
1988), and glass (MacLeod & Davies, 1987). A. Elena Charola (2000) gives a broad
overview of the mechanisms of salt efflorescence and the damage it causes in all affected
materials.

1.3.1 Desalination and Ceramics

Identifying the soluble and insoluble salts present in efflorescence on ceramics
can help to control and inform the desalination process (Odegaard, et al., 2011; Paterakis,
1987). Different soluble salts require longer to diffuse into wash water or poultices based
on their crystalline structure (Koob & Ng, 2000). Ceramics are usually desalinated by

6
washing in running water, soaking in a bath of deionised water, or using a poultice
(Odegaard, et al., 2000).
The extent of efflorescence can determine how salts are removed. The most
common method of desalination involves soaking the object in deionised water until the
levels on a conductivity meter reach an acceptable level. There are no standards for what
these levels are and many authors focus on the negative effects of desalination by
soaking, such as softening of the ceramic and the removal of archaeological evidence
(Johnson, 1998; Ling & Smith, 1996; Unruh, 2001). Past archaeological good practice
usually promulgated soaking objects for weeks to remove salts, but more recent papers
advise a maximum of 5 days or even 24 hours is enough to remove soluble salts from
objects (Unruh, 2001; White, et al., 2010). If the surface is friable, the ceramic can be
consolidated in part or in full prior to desalination with Paraloid B-72 and other
adhesives, or with cyclododecane (Ellis, et al., 2007; Muros, V. & Hirx, J. 2004).
Beyond potentially damaging the ceramic, desalination also removes evidence of
burial and residues from use (Koob & Ng, 2000; Ling & Smith, 1996; Unruh, 2001;
Willey, 1995). However, many authors agree that desalination of actively spalling
ceramics prevents future damage to the artefacts (Koob & Ng, 2000; Ling & Smith,
1996; Unruh, 2001).

1.3.2 Commonly Used Instrumentation

Archaeologists and conservators use a variety of methods to test for soluble
salts—before and after desalination treatment. These methods include: wet chemical spot
tests; optical microscopy, polarised light microscopy, portable X-ray fluorescence, X-ray
diffraction, scanning electron microscopy with energy-dispersion X-ray spectroscopy,
ionic chromatography, and hygroscopic moisture content analysis.

1. Wet Chemical Spot Tests. Probably the simplest way to test for salts involves
the use of wet chemical spot tests. Wet chemical spot tests tell what salts are present in
salt solutions, usually in the form of precipitates (Koob & Ng 2000; Odegaard, et al.
2000; Odegaard, et al. 2011). Some tests such the test of carbonates using barium
hydroxide and test for phosphates (Odegaard, et al. 2001, 102:120) were very difficult to
interpret. Conclusive molecular identification is therefore hard to obtain. Some have also
tested the use of chemical indicator test strips (Ellis, et al., 2007; Frame, et al., 2007;
Odegaard, et al., 2011). Ellis, Derrick, and Newman (2007) posit that commercially

7
available indicator strips for sulphates and nitrates only accurately identify the salts in
concentrated wash waters—not with samples collected with cotton swabs. However,
they conclude that chemical spot tests, while more accurate than test strips, do not
provide a measure of the levels of salts in the ceramic (Ellis, et al., 2007, 69).

2. Visual Identification. Scientists can classify soluble salts visually if they present
with large enough crystals through optical microscopy. The size and shape of crystals can
be used to identify individual mineral species (Nesse, 2000). Optical microscopy can be
used to identify larger crystals, while polarised light microscopy (PLM) works well for
smaller crystals (Odegaard, et al., 2011, 14).

3. pXRF. Portable X-ray fluorescence (pXRF) provides a non-destructive way to analyse

which elements are present in a sample. They can be used directly on the surface of
ceramics prior to desalination (Gazulla, et al., 2010; Jin, et al., 2014). However, this
method only detects elements heavier than magnesium. The pXRF also has a limited
ability to detect low-Z orbitals of lighter elements. This becomes a problem because it
would be unable to detect sodium. If UC18699 was used for offerings or held natron, it
would probably have sodium present, either as Na2CO•10H2O (sodium carbonate
decahydrate, ie. natron) or NaCl (sodium chloride). Another problem is that the findings
from a pXRF are limited to percentages or parts per million. Though it is not a perfect
solution, the pXRF represents a qualitative way to determine that certain elements are
present in a sample. The results obtained by pXRF should not be interpreted as accurate,
quantitative data on the actual percentage of each element in the ceramic (Gazulla, et al.,
2010; Jin, et al., 2014).

4. XRD. X-ray diffraction (XRD) easily identifies crystals from small sample sizes, a
useful tool in determining soluble salts prior to desalination (Ellis, et al., 2007; Gazulla, et
al., 2010; Jin, et al., 2014: Odegaard, et al., 2011; Wheeler & Wypyski 1993). Data is
presented as a series of peaks. Each peak represents a unique identifier for different
crystal structures. The XRD does not easily identify mixtures of material. This is
especially problematic when trying to identify many types of salts in a sample that is
mixed with pottery. Others have corrected for this by using wash water and evaporating
out the salt crystals (Bradley & Middleton, 1988). Other forms of analysis, such as PLM
and SEM-EDS will help corroborate results from data that is hard to interpret.

8
5. SEM-EDS. Scanning electron microscopy with energy-dispersive X-ray spectroscopy
(SEM-EDS) can be used to classify salts visually by their crystalline structure (Jin, et al.,
2014; Wheeler & Wypyski, 1993). The EDS allows for the identification of elements
within the sample. Data is collected with a collimated beam by selecting specific points
on the SEM image. Many conservators and museums may not have access to this costly
and time-consuming technology. It can be used to identify mixtures by identifying
individual crystals in a sample. It is useful in conjunction with data from the XRD and
other analytical methods.

6. Other Methods. Conservators also employ other methods to identify salts.

Conservators at the Arizona State Museum (ASM) have developed a prototype for a
portable tool for testing soluble salts, non-destructive on the surface of ceramics to give
quantitative results (Frame, et al., 2007). Other methods include ionic chromatography
(Gazulla, et al., 2010; Jin, et al., 2014; Odegaard, et al., 2011; Pessoa, et al., 1996) and
analysing the hygroscopic moisture content (HMC) of salts in ceramics (Gonçalves, et al.,
2006). Ionic chromatography and the ASM prototype are not available at UCL, while
HMC requires the use of large sample sizes to identify salts (Gonçalves, et al., 2006).

1.4 Objectives
This study seeks to identify the composition of soluble and insoluble salts, which
have effloresced on the surface of UC18699. Analytical techniques utilized will include
wet chemical spot tests, microscopy (optical and PLM), pXRF, and SEM-EDS. These
will be used as supplementary material for the main study, using the XRD. All of these
methods have been identified by pervious studies as useful in identifying salts (see 1.3.2).
An experiment will then be conducted using known salts mixed with low-fire ceramic to
attempt to replicate the spalled sample under XRD. This experiment will aid in the
identification of the UC18699 sample as well as attempt to determine the composition of
salts in a mixture.

1.5 Relevance and Importance

As previously mentioned, most studies of soluble salts utilize concentrated wash
water from desalination or sample material from the surface of pottery. This study hopes
to use a sample of spalled material to obtain the composition of salts in UC18699.

9
Knowing the chemical composition of the soluble salts as well as what parts of the object
have insoluble accretions will aid in tailoring the desalination process to UC18699’s
needs.
Analogous material made from crushed terra cotta pots mixed with known
mixtures of salts was used to help interpret ambiguous results under the XRD. This will
be used to show if spalled material can be used to determine salt composition before
desalination without taking samples from the friable surface of the object. Not all
conservators and museums have access to the materials and equipment used in these
analyses. This paper will hopefully show that the results of an in-depth analysis of one
representative object from a collection (such as the Harageh collection at the Petrie
Museum) can be used to more confidently identify salt efflorescence on other objects
using less technological means, such as wet chemical spot tests.

10
II. Materials and Methods
Analysis of UC18699 and its soluble salts took place at the University College London,
Institute of Archaeology’s Conservation Laboratory, Conservation Research Laboratory,
and the Wolfson Archaeological Science Laboratories. Only non-destructive techniques
wereused. Wet chemical spot tests, optical microscopy, pXRF, XRD, and SEM-EDS
have been used to analyse the unknown salt and pottery mixture. A more extensive
experiment with terra cotta pots and salts was conducted with the XRD. All analytical
techniques used the spalled material collected from UC18699’s packaging material.
Instrumentation and Models
Instrument Specifications
Digital Dino-Lite Pro HR AM-7013MT; Dino-Lite Pro
MicroscopeS AD4113T-FVW Ultra-violet Digital Microscope
pXRF Olympus Delta Premium DP-4000
XRD Rigaku Miniflex 600
SEM-EDS Hitachi SEM S-3400 N fitted with an Oxford
Instruments EDS 7021 INCA X-ray Energy
Dispersive Spectrometer
Table 1. Instrumentation specifications to be utilized in this study.
Analyses Performed
1. Visual Analysis: Optical Microscopy, UV Light, UV Dino Light
The object was first observed under a Dino-Lite Pro HR AM-7013MT Digital
Microscope and a Dino-Lite Pro AD4113T-FVW Ultra-violet Digital Microscope. This
magnification allowed some basic crystalline structure to be observed. UC18699 was also
observed under an ultraviolet light. This served to see if any material on the jar, such as
adhesive fluoresced in the presence of UV light.

2. Portable X-ray Fluorescence

The next step in salt identification was performed using a handheld Olympus Delta
Premium DP-4000 pXRF in Mining Mode (Two-Beam) with a gold (Au) tube anode.
The device was programmed for a 120 second live time (60 seconds per beam). The spot
size of each beam was 70-100 µm. Beam 1 measures lighter elements, while Beam 2
measures heavier elements. Current and voltage were between 15-40 kV. Results were
reported as a percentage rather than parts per million. Tests were conducted directly on
the surface of the ceramic; no sample preparation was necessary. Though it identifies
elements present in the sample, it does not provide precise molecular identification. It
was used a first step to determine which wet chemical spot tests to perform based on
elements present in three distinct areas.

11
3. Wet Chemical Spot Tests
Standard wet chemical spot tests were performed for soluble salts: chlorides (silver
nitrate test), nitrates, and sulphates; insoluble salts: carbonates ((1)simple nitric acid, and
(2) barium hydroxide) and phosphates. Methods for the wet chemical spot tests were
taken from: Buys & Oakley, 1993; Odegaard, et al., 2001; Odegaard, et al., 2011; Riss,
1993.
Salts Tested with Wet Chemical Spot Tests
Soluble Test Performed Insoluble Test Performed
Salts Salts
Chlorides Silver nitrate test Carbonates Nitric acid test
Nitrates Iron (II) sulphate test Carbonates Barium hydroxide and
hydrochloric acid test
Sulphates Barium chloride test Phosphates Ammonium molybdate
and ascorbic acid
Table 2. Table adapted from Johnson, 1998, 2. Spot tests from Odegaard et al., 2001.

4. X-ray Diffraction
A small sample of the salt and pottery mixture was analysed with an XRD, a Rigaku
Miniflex 600. Tube voltage was set to 40kV and the current was set to 15 amps. The start
degree was 3 and stop degree was 90. Speed was 4 with a step of 0.020. Sample
preparation involved grinding a small amount of the mixture with a mortar and pestle.
The fine powder was placed on a silicon-alloy sample tray and placed in the machine for
analysis.

5. SEM-EDS
The spalled sample was also imaged in secondary electron mode under the SEM-EDS
(Hitachi SEM S-3400 N fitted with an Oxford Instruments EDS 7021 INCA X-ray
Energy Dispersive Spectrometer). The SEM allows the researcher to more easily and
accurately determine crystalline structure in fine detail through visual identification and
elemental identification when used in combination with EDS. Optimisation of the EDS
was performed with a cobalt sample. The sample of salts was fixed to an SEM stub with
double-sided carbon tape. Imaging was taken at S3400 at 20.0KV. Images for analysis
under the EDS were taken at 10.0 mm x 200 SE. Point analysis under the EDS revealed
the elements present in individual crystals of the sample. Area analysis was also used to
show the range of salts in the sample. SEM imaging and EDS allowed for the definitive
identification of different mineral crystals, especially those unable to be confirmed with
other methods.

12
6. XRD Experimental Procedure
An experiment was conducted using the XRD to try and determine the composition of a
mixture of salts and pottery. It was performed to help interpret the inconclusive results
of the XRD obtained by sampling the spalled material. Eight tests in total were
conducted using known salts (sodium chloride, sodium sulphate (anhydrous), and
calcium carbonate) and pottery based on the information obtained from the testing on
the spalled sample. Each of the salts and ceramic were isolated to make them more easily
identified under the XRD in samples 1-4. Samples 5-7 contain mixtures of one salt with
the ceramic, in a 1:1 ratio. In sample 8 all of the salts were mixed with the ceramic in a
1:1:1:1 ratio. Potential problems with this method could include overlaps in peaks on the
XRD. This would especially render the mixed sample difficult to interpret.
Experimental Mixtures
Sample Mixture
1 Terra cotta
2 Calcium
carbonate
3 Sodium chloride
4 Sodium sulphate
(anhydrous)
5 Terra cotta Calcium
carbonate
6 Terra cotta Sodium chloride
7 Terra cotta Sodium sulphate
(anhydrous)
8 Terra cotta Calcium Sodium Sodium sulphate
carbonate chloride (anhydrous)
Table 3. Experimental mixtures to be run through the XRD.

13
III. Results
1. Visual Analysis
1.1 Digital Microscope

Viewing the surface of the pottery under the Dino-Lite produced the following images.
The magnification was too weak for crystals to be identified. However, the Dino-Lite
elucidated the extent of the friable surface, cracks, and pits from spalling.

Dino Light Photos

Detail of grey and white crystals inside one of the Detail of small hole with grey and white crystals,
cracks on the rim white crystals on the surface

Detail of grey crystals in a hairline crack, white Small hole left from spalling, grey and white crystals
crystals on the surface on surface and in void
Table 4. Images of UC18699 obtained using the Dino-Lite.

14
1.2 UV Light/UV Dino Light
No identifiable features such as adhesive
were observed under the UV light and UV
Dino-Lite. The only fluorescing material
consisted of white dots throughout the
ceramic body (Figure 5, right). These could
potentially be a species of salt fluorescing or
carbonates in the body of the ceramic from
the clay used to make UC18699. Figure 6. UV Dino-Lite shows small fluorescent
dots in the surface of UC18699

2. Spot Tests
Results Positive/Negative
1
Chlorides White precipitate (AgCl) observed Positive for chlorides
Carbonates (1)2 Bubbling observed Inconclusive for carbonates
Carbonates (2)3 Unable to discern cloudy precipitate Inconclusive for carbonates
when Ba(OH)2 suspended over gas
from previous test
Sulphates4 White precipitate observed Positive for sulphates
Nitrates5 No observable reaction Negative
6
Phosphates No observable reaction Negative
Table 5. Results of spot tests performed on spalled material. Followed guidance of Odegaard et al. 2000.

1 Sample placed in a watch glass over black paper to aid in visibility. 2 drops 1M nitric acid (HNO3) added.
2 drops 2M silver nitrate added. White precipitate (AgCl) observed.
2 Sample placed in a watch glass over black paper to aid in visibility. 2 drops 1M nitric acid (HNO ) added.
3
Bubbling observed.
3 Solution of Ba(OH) suspended over gas from original carbonate test (above) in a test tube with the help
2
of Dr. Caitlin O’Grady. Unable to discern cloudy precipitate.
4 Sample placed in a watch glass over black paper to aid in visibility. 2 drops 1M nitric acid (HNO ) added.
3
2 drops of 0.5M barium chloride (BaCl2) added to the solution. Sulphates are present, white precipitate
(BaSO4) observed.
5 Sample placed in a watch glass over white paper to aid in visibility. 2 drops solution of diphenylamine

((C6H5)2NH) and ammonium chloride (NH4Cl) in sulphuric acid (H2SO4) added (UCL lab solution). No
blue precipitate observed.
6 Sample placed in a watch glass over black paper to aid in visibility. 2 drops 1M nitric acid (HNO ) added.
3
1 drop 0.5M ammonium molybdate added. No yellow precipitate observed.

15
3. Portable X-ray Fluorescence
3.1. Location 1: The pXRF recorded a high percentage elemental composition of
calcium (Ca) in the areas of the jar with white surface deposits. There was also a rather
high percentage of sulphur (S) in this
location.
Location 1: Major and Minor
Elements
Major Minor
Elements Elements
Calcium (Ca) Aluminium (Al)
Chlorine (Cl) Iron (Fe)
Potassium (K) Silicon (Si)
Sulphur (S)
Table 6. Showing the major and minor
elements, Location 1.

3.2 Location 2: The pXRF

recorded a high percentage elemental
composition of potassium (K),
silicon (Si), and aluminium (Al).
These percentages are consistent
with ceramic composition.
Location 2: Major and Minor
Elements
Major Minor
Elements Elements
Aluminium (Al) Calcium (Ca)
Iron (Fe) Chlorine (Cl)
Potassium (K) Sulphur (S)
Figure 7. Picture shows where pXRF tests were performed Silicon (Si)
on UC18699. Location 1: White surface deposits. Location Table 7. Showing the major and minor
2: Area of the surface without visible salt efflorescence. elements, Location 1.
Location 3: Gray surface deposits.
3.3 Location 3: The pXRF
recorded a high percentage elemental composition of chlorine (Cl) in the areas of the jar
with grey surface deposits.
Location 3: Major and Minor Elements
Major Elements Minor Elements
Chlorine (Cl) Aluminium (Al) Potassium (K)
Calcium (Ca) Silicon (Si)
Iron (Fe) Sulphur (S)
Table 8. Showing the major and minor elements, Location 3.

16
 4. X-ray Diffraction
4.1 UC18699
Analysis using the XRD’s software matched the diffraction patterns of the small sample
to databases of known materials. The results showed the presence of sodium chloride
(NaCl) and silica (SiO2).

8000

6000

4000
Intensity (cps)

2000

0
Sodium Chloride, Na Cl, 01- 078- 0751
Quartz alpha, Si O2, 01- 085- 0865
20 40 60 80
2-theta (deg)

Figure 8. XRD spectra of small mixed sample, pottery and soluble salts. Shows likely presence of sodium chloride, NaCl
(blue) and quartz silica, SiO2 (red).

4.2 Results of Analogous Experiment

Analysis using the XRD’s software was performed on eight samples of known salts and
pottery. The diffraction patterns of the samples in known ratios were compared to those
of UC18699 using the JMP Software. This was an attempt to match the crystal structures
in UC18699 that did not match diffraction patterns in the ICDD software.

17
 2500

2000

1500

1000
Intensity (counts)

500

0
20 40 60 80
2-theta (deg)

Figure 9. XRD spectra of known sodium chloride and terra cotta mixed sample. Salt and pottery were
mixed in a 1:1 ratio by weight.

500

400

300
Intensity (counts)

200

100

20 40 60 80
2-theta (deg)

Figure 10. XRD spectra of known sodium chloride, calcium carbonate, sodium sulphate (anhydrous), and
terra cotta sample. Salts and pottery were mixed in a 1:1:1:1 ratio by weight.

18
 1000

800

600
Intensity (counts)

400

200

0
20 40 60 80
2-theta (deg)

Figure 11. XRD spectra of known sodium chloride, calcium carbonate, sodium sulphate (anhydrous), and
terra cotta spectra (blue) overlaid with UC18699 spectra (red). Salts and pottery were mixed in a 1:1:1:1
ratio by weight.

19
5. SEM-EDS
5.1 Site of Interest 1: Analysis of three areas on the salt sample produced three
distinct spectra. Particle 1 was chosen because of its sharp edges in contrast to particle 2,
which is more round. Both particles are roughly the same size. Spectrum 3 was chosen
because of its small particle size. Analysis of the total area was undertaken because of the
great variety in particle shape and size. Spectrum 4 showed a wide variety of elements.
Spectra Major Elements Minor Elements

1 Chlorine (Cl) Boron (B)

Sodium (Na) Calcium (Ca)
2 Silicon (Si) Calcium (Ca) Potassium (K)
Oxygen (O) Iron (Fe) Titanium (Ti)
Potassium (K)
Titanium (Ti)
3 Calcium (Ca) Chlorine (Cl)
Potassium (K) Phosphorus (P)
Sulphur (S)
4 Aluminium (Al) Iron (Fe) Sulphur (S)
(Area Calcium (Ca) Magnesium (Mg) Phosphorus (P)
Spectrum) Chlorine (Cl) Potassium (K)
Silicon (Si) Sodium (Na)
Table 9. Results of the elemental analysis under the EDS shows the major and minor elements in each area
of analysis. Full spectra in Appendix 4.1, Figure 6.

20
5.2 Site of Interest 2: Analysis of three areas on the salt sample produced three
distinct spectra. Particle 1 was chosen because of its sharp edges in contrast to particle 2,
which is more round and spongy in appearance. Particle 1 is also much larger than
particle 2. Particle 3 is smaller, smoother, and sharper looking. Analysis of the total area
was undertaken because of the great variety in particle shape and size. Spectrum 4
showed a wide variety of elements.

Spectra Major Elements Minor Elements

1 Chlorine (Cl) Aluminium (Al) Silicon (Si)
Sodium (Na) Calcium (Ca) Sulphur (S)
2 Calcium (Ca) Calcium (Ca) Potassium (K)
Carbon (C) Iron (Fe) Sulphur (S)
Oxygen (O) Magnesium (Mg)
Silicon (Si)
3 Chlorine (Cl) Aluminium (Al) Oxygen (O)
Sodium (Na) Calcium (Ca) Silicon (Si)
4 Aluminium (Al) Iron (Fe) Sulphur (S)
(Area Calcium (Ca) Magnesium (Mg) Phosphorus (P)
Spectrum) Chlorine (Cl) Potassium (K) Titanium (Ti)
Silicon (Si)
Sodium (Na)
Table 10. Results of the elemental analysis under the EDS shows the major and minor elements in each
area of analysis. Full spectra in Appendix 4.2, Figure 7.

21
 Figure 12. SEM images of spalled sample. Top image shows the heterogeneity of the sample and the great
variety in particle size and shape. Bottom image shows a close-up view of pottery (flat, plate-like particles)
interspersed with small, greyer, irregular salt crystals.

22
IV. Discussion
According to the completed analyses, elements and compounds found in the spalled
material from UC18699 include: aluminium, calcium, chlorine, carbonate, iron,
potassium, silicon, and sulphur (Table 9).

Elements in UC18699
Corroborating Al Ca Cl CO3 Fe K Si S
Instrumentation
pXRF X X X X X X X
Wet Chemical X I X (SO4)
Spot Tests
XRD X X
(NaCl) (SiO2)
SEM-EDS X X X I X X X X
Table 11. Shows elements present in UC18699 as found by completed analyses (X). Presence of carbonates
is marked (I) for inconclusive.

Aluminium and silica (SO2) form the basic components of clay. Iron is also
commonly found as an impurity in clays, but it could also potentially be bonded with one
of the anions present in UC18699 as part of a salt (Cronym, 1990: 145).
The salts most commonly found in efflorescing ceramics include soluble salts:
chlorides, nitrates, and sulphates; and insoluble salts: carbonates and phosphates
(Johnson 1998). Three of these common types were found in the spalled sample by
multiple analytical tests: sodium chloride (NaCl) and an unidentified sulphate and
carbonate. The SEM-EDS also picked up small amounts of phosphorus. This could be
the result of phosphorus in the pottery from the firing process or from the efflorescence
of insoluble salts from burial.
Results for the carbonate tests were inconclusive. In the second, more involved
test it was impossible to tell if the aqueous solution of Ba(OH)2 was cloudy from carbon
dioxide gas or from a contaminated sample. Because the test relies on the boiling of
distilled water to remove CO2, it can easily become contaminated and turn cloudy by
being exposed to the air. The wet chemical lab facilities used in the Conservation Lab
were not ideal for carrying out this test. SEM-EDS seems to point to the presence of
calcium carbonate. However, carbon and oxygen are not necessarily present in the
sample in carbonate form.
Though the tests performed give a general idea of what soluble salts might be
present (sodium chloride, carbonates, and sulphates) results are inconclusive. One reason
for this could have been the use of only a spalled sample. If insoluble salts such as
carbonates and phosphates were present in the ceramic jar, they would not necessarily

23
have spalled off the surface in large amounts. Unlike soluble salts, insoluble salt
efflorescence does not cause catastrophic damage to ceramics and other porous
materials.
The results of the XRD experiment showed that mixed salt and pottery samples
could be easily identified. However, the mixed samples tested in the experiment were
known mixtures in 1:1 ratios. When analysing the samples in ICDD, some of the salts
were misidentified. If the researcher did not know what salts were present in the sample
prior to testing them in the XRD, this would have complicated matters greatly. The
XRD can be used to identify mixtures, but it should be used in combination with a
variety of methods. This way conservators can have an idea of what elements and salts
might be present in the sample to aid in identification of peaks.

24
V. Conclusions
Insoluble and soluble salts were both present in the spalled sample from UC18699.
These included sodium chloride, carbonates, and sulphates. After identifying soluble salts
in the sample, the decision was made to desalinate UC18699.
The ceramic was suspended
in a Stewart box in a nylon
gossamer sling, held to the
box with zip ties. The jar
was soaked for three days,
until the conductivity metre
reading had not changed in
24 hours.
The desalination process
removes evidence of use and
burial environment. It also
stabilises ceramics and other
porous materials by
removing an internal agent
of deterioration (salts).
Desalination is a process
which endangers porous
materials. Conservators need
to be sure that there are Figure 13. Photos of the desalination process. Top: UC18699 is fully
submerged in deionized water. Bubbles forming on the surface show
soluble salts in the ceramic the movement of air and salts out of the ceramic body. Bottom: the
desalination set-up. The white material covering the letters and
prior to desalination in order vulnerable areas is cyclododecane, a waxy substance that sublimes at
room temperature.
to decide how to proceed.
As seen in Figure 5, sulphates are less soluble than chlorides. An artefact, which has
tested positive for sulphates, should potentially be soaked longer than one with only
nitrates to make sure all the salts have migrated out of the ceramic body.
Not all conservation departments or archaeological sites have access to the sophisticated
instrumentation utilised in this study. However, no one method (optical microscopy, wet
chemical spot tests, pXRF, XRD, or SEM-EDS) definitively confirms the presents of
salts on its own. Even in combination, the results of the tests are inconclusive. The most

25
accurate and cost-effective method for museums and conservators to use are still
probably wet chemical spot tests.

Recommendations for Future Study

This study had hoped to conclude that there was a
definitive way to identify salts using spalled
samples. Non-destructive methods identified
broadly the elements present in the sample.
Destructive methods of analysis such as thin-
sectioning the pottery could have allowed for a
more definitive view of how the salts had
crystallised within the ceramic body. As previously
discussed (Section 1.3.2, 6) ionic chromatography
and hygroscopic moisture content could be tested
to see how accurately they can identify salts from
a spalled sample. These methods were not
available for this study.

Figure 14. UC18699 after desalination.

26
VI. Acknowledgements
I would like to especially thank Dr. Caitlin O’Grady, Agnese Benzonelli, Dr. John
Merkel, James Hales, Dean Sully, Tom Gregory, and Patrick Quinn for all of their help
and support in completing this project.

VII. References
Abdel-Maksoud, G. and El-Amin, A-R. 2011. A Review on the Materials Used During the
Mummification Processes in Ancient Egypt. Mediterranean Archaeology and Archaeometry
11/2, pp. 129-150.
Bradley, S.M. & Middleton, A.P. 1988. A Study of the Deterioration of Egyptian Limestone
Sculpture. Journal of the American Institute for Conservation 26/2 (Autumn), pp. 64- 86.
Buys, S. & Oakley, V. 1993. The Conservation and Restoration of Ceramics, Oxford:
Butterworth-Heinemann.
Charola, A.E. and Herodotus 2000. Salts in the Deterioration of Porous Materials: An
Overview. Journal of the American Institute for Conservation Online 39/3:2 [accessed 01
March 2015]. Available at: < http://cool.conservation- us.org/jaic/articles/jaic39-03-
002.html>
Cronym, J.M. 1990. The Elements of Archaeological Conservation, London: Routledge.
De Alarcon, T., 2013. The Salty Pots of Ur and the Desalination Station. Penn Museum Blog.
[accessed 19 November 2014]. Available at: <http://www.penn.museum/blog/
collection/conservation/the-salty-pots-of-ur-and-the-desalination-station/>
Ellis, L., Derrick, M., & Newman, R. 2007. In: L. Pilosi, ed. Glass and Ceramics
Conservation 2007: Interim Meeting of the ICOM-CC Working Group. Nova Gorica : Goriški
muzej, pp. 62-71.
Engelbach, R. 1923. Harageh, Egypt: British School of Archaeology in Egypt.
Frame, L.D., Segalman, I., & White, C. 2007. Conductivity probe prototype: measuring
soluble salt content on ancient ceramics. WAAC Newsletter 29/1, pp. 12-14 [accessed 19
November 2014]. Available at: <http://cool.conservation-us.org/waac/wn/wn29/
wn29-1/wn29-105.pdf>
Gazulla, M.F., Gómez-Tena, M.P., Orduna, M., Vincente, S., & Amorós, J.L. 2010.
Determinación de sales solubles y eflorescencias en tejas cerámicas. Boletín de la Sociedad
Española de Cerámica y Vidrio 49/3, pp. 189-196 [in Spanish].
Goncalves, T.D., Rodrigues, J.D., & Abreu, M.M. 2006. Evaluating the salt content of salt-
contaminated samples on the basis of their hygroscopic behavior: Part II: experiments
with nine common soluble salts. Journal of Cultural Heritage 7/3, pp. 193-200.
Jin, P., Zhang, W., Wang, Q., Yang, X., Sun, S., Fan, X., & Li, B. 2014. Research Into Water-
Soluble Salts in Efflorescent Pottery During Long-Term Storage in a Museum. Corrosion
Science 89, pp. 268-274.
Johnson, J. 1998. Soluble Salts and Deterioration of Archaeological Materials. Conserve O
Gram, National Park Service 6/5, pp. 1-4. [accessed 12 November 2014]. Available at:
<http://www.nps.gov/museum/publications/conserveogram/06-05.pdf>
Koob, S.P. & Ng, W.Y. 2000. The Desalination of Ceramics Using a Semi-Automated
Continuous Washing Station. Studies in Conservation 45/4, pp. 256-273.
Ling, D. & Smith, S. 1996. To desalinate or not to desalinate? That is the question. In Le
Dessalement des Matériaux Poreux, 7éme Journées d’études de la SFIIC, Poitiers. Paris: SFIIC,
pp. 65-74.
Muros, V. & Hirx, J. 2004. The Use of Cyclododecane as a Temporary Barrier for Water-
Sensitive Ink on Archaeological Ceramics during Desalination. Journal of the American
Institute for Conservation 43/1 (Spring), pp. 75-89.
Nesse, W. 2000. Introduction to Mineralogy. Oxford: Oxford University Press.
Odegaard, N., Carroll, S., & Zimmt, W.S. 2000. Material Characterization Tests: for Objects
of Art and Archaeology (Second Edition). London: Archetype Publications.
Odegaard, N., Hill, P., Santarelli, B., & Zimmt, W. 2011. Detecting and Identifying Salts

27
 during the Desalination Process with Spot Test Papers. WAAC Newsletter 33/1
(January), pp. 14-17 [accessed 16 October 2014]. Available at:
<http://cool.conservation-us.org/waac/wn/wn33/wn33-1/wn33-104.pdf>
Pancaldo, S. 2014. 15 October. [Personal communication]. London: Institute of Archaeology.
Paterakis, A.B. 1987. The Deterioration of Ceramics by Soluble Salts and Methods for
Monitoring their Removal. In J. Black, ed. Recent Advances in the Conservation and Analysis of
Artefact. London: Summer Schools Press.
Paterakis, A.B. 1998. The Desalination of Consolidated Ceramics. In A.B. Paterakis,
ed. Glass, Ceramics and Related Materials: Interim Meeting of the ICOM-CC Working Group
September 13-16, 1998 Vantaa, Finland., Vantaa, Finland: EVTEK-Institute of Arts and
Design, pp. 144-153.
Pessoa, J.C., Farinha Antunes, J.L., & Fortes, M.A. 1996. Removal and Analysis of Soluble
Salts from Ancient Tiles. Studies in Conservation 41/3, pp. 153-160.
Petrie Museum of Egyptian Archaeology 2014. Collections Online. Petrie Museum of
Egyptian Archaeology [accessed 15 October 2014]. Available at: <petriecat.
museums.ucl.ac.uk>
Riss, D. 1993. Testing for chlorides with silver nitrate. Conserve-O-Gram 6/3. National Park
Service [accessed 24 October 2014]. Available at: <http://www.nps.gov/museum/
publications/conserveogram/06-03.pdf>
Universiy College London 2001. Harageh (Haraga). Digital Egypt for Universities [accessed
17 October 2014]. Available at: <http://www.digitalegypt.ucl.ac.uk/harageh/index.
html>
Unruh, J. 2001. A Revised Endpoint for Ceramics Desalination at the Archaeological Site of
Gordion, Turkey. Studies in Conservation 46/2, pp. 81-92.
Wheeler, G.S. & Wypyski, M.T. 1993. An Unusual Efflorescence on Greek Ceramics. Studies
in Conservation 38/1 (February), pp. 55-62.
White, C., Pool, M., & Carroll, N. 2010. A Revised Method to Calculate Desalination Rates
and Improve Data Resolution. Journal of the American Institute for Conservation 49/1
(Spring/Summer), pp. 45-52.

28
VIII. Appendix

1. Image of UC18698 and UC18699 in Engelbach, 1923

Figure 1. Engelbach, 1923, pp. 10; Plate XXXIX, Figures 67O (UC18698) and 67O2 (UC18699).

29
2. pXRF Data of Major and Minor Elements

Element Percentage (%) Standard

Deviation
Al 3.30 0.69
Si 8.04 0.22
S 5.74 0.04
Cl 27.03 0.28
K 15.77 0.12
Ca 12.99 0.03
Fe 5.13 0.02
Light elements 31.24 0.30
Table 1: Location 1: high percentage composition of calcium (Ca) in the areas of the jar with white surface
deposits. Also a rather high percentage of sulphur (S).

Element Percentage (%) Standard

Deviation
Al 6.88 0.48
Si 17.79 0.21
S 1.25 0.02
Cl 14.46 0.19
K 26.01 0.10
Ca 6.13 0.02
Fe 6.85 0.02
Light elements 19.26 0.13
Table 2: Location 2: percentage elemental composition of an area of the jar without visible surface
deposits. Shows high percentage of potassium (K), silicon (Si), and aluminium (Al). These percentages are
consistent with ceramic composition.

Element Percentage (%) Standard

Deviation
Al 2.52 0.35
Si 6.65 0.11
S 3.36 0.02
Cl 38.52 0.20
K 11.33 0.07
Ca 9.80 0.02
Fe 4.19 0.01
Light elements 22.86 0.16
Table 3. High percentage composition of chlorine (Cl) in the areas of the jar with grey surface deposits.

30
3. XRD Qualitative Analysis Result Reports

3.1 Qualitative Analysis Results: Spalled Sample from UC18699

General information
Analysis date 2014/11/21 15:16:38
Sample name UC18699 mixed Measurement date 2014/11/21 14:59:47
pottery/salts
File name SAHCeramic_Theta_2- Operator
Theta.raw
Comment Full Chart

Qualitative analysis results

Phase name Formula Figure of merit Phase reg. detail DB card number
Sodium Chloride Na Cl 0.561 ICDD (PDF2.DAT) 01-078-0751
Quartz alpha Si O2 1.008 ICDD (PDF2.DAT) 01-085-0865

Phase data pattern

8000

6000

4000
Intensity (cps)

2000

0
Sodium Chloride, Na Cl, 01- 078- 0751
Quartz alpha, Si O2, 01- 085- 0865
20 40 60 80
2-theta (deg)

Figure 2. Data from the XRD. Shows diffraction pattern matches for silica and sodium chloride.

31
3.2 Qualitative Analysis Results: 1:1 Sodium Chloride and Terra
Cotta Mixed Sample (RIR)

General information

Analysis date 2015/03/16 16:06:32
Sample name sample6 Measurement date 2015/03/16 15:38:41
File name Maddiesample6_Theta_2- Operator
Theta.raw

Qualitative analysis results

Phase name Formula Figure of merit Phase reg. detail DB card number
Halite, syn Na Cl 0.311 ICDD (PDF2.DAT) 00-005-0628
Silicon Oxide Si O2 0.520 ICDD (PDF2.DAT) 01-085-0794

Weight ratio

Phase name Content(%)
Halite, syn 53.5(5)
Silicon Oxide 46.5(6)

Maddiesample6_Theta_2-Theta

0 20 40 60 80

Wt(%)

Halite, syn Silicon Oxide

Unknown

Measurement profile

2500

2000

1500

1000
Intensity (counts)

500

0
20 40 60 80

2-theta (deg)

Figure 3. Data from the XRD. Shows diffraction pattern matches for silica and sodium chloride.

32
3.3 Qualitative Analysis Results: 1:1:1:1 Sodium Chloride,
Calcium Carbonate, Sodium Sulphate (Anhydrous) and Terra
Cotta Mixed Sample Overlaid with UC18699 Spalled Sample
General information
Analysis date 2015/03/18 15:04:52
Sample name UC18699 Measurement date 2014/11/21 14:59:47
File name SAHCeramic_Theta_2- Operator
Theta.raw
Comment Full Chart

Measurement profile


1000

800

600


Intensity (counts)

400


200





0
20 40 60 80

2-theta (deg)

Figure 4. Data from the XRD.

33
3.4 Qualitative Analysis Results: 1:1:1:1 Sodium Chloride,
Calcium Carbonate, Sodium Sulphate (Anhydrous) and Terra
Cotta Mixed Sample (RIR)
General information
Analysis date 2015/03/18 15:08:54
Sample name sample8 Measurement date 2015/03/18 14:39:31
File name Maddiesample8_Theta_2- Operator
Theta.raw
Comment

Qualitative analysis results

Phase name Formula Figure of merit Phase reg. detail DB card number

Measurement profile

500

400

300
Intensity (counts)

200

100

20 40 60 80
2-theta (deg)

Figure 5. Data from the XRD. Shows diffraction pattern matches for silica and sodium chloride.

34
4. SEM Spectra
4.1 Area of Interest 1, Spectra 1-3
Spectrum 1

Spectrum 2

Spectrum 3

Figure 6. SEM Results, area of interest 1, spectra 1-3. Spectrum 1 shows high percentage NaCl. Spectrum 2 shows
a high percentage of SiO2. Spectrum 3 shows a high percentage of Ca, K, and O2.
35
4.2 Area of Interest 2, Spectra 1-3
Spectrum 1

Spectrum 2

Spectrum 3

Figure 7. SEM Results, area of interest 2, spectra 1-3. Spectrum 1 shows high percentage NaCl. Spectrum 2 shows
a high percentage of Ca. Spectrum 3 shows a high percentage of NaCl as well
36

4.3 Area of Interest 1, Spectrum 4: Area Analysis

Figure 8. SEM Results, area of interest 1, spectrum 4. Spectrum 4 shows high percentage SiO2, Ca, NaCl,
Al, and S.

4.4 Area of Interest 2, Spectrum 4: Area Analysis

Figure 9. SEM Results, area of interest 2, spectrum 4. Spectrum 4 also shows high percentage SiO2, Ca, NaCl, Al,
and S.

37