Citation
The Methodological development of lipid residue analysis in archaeological pottery samples

Material Information

Title:
The Methodological development of lipid residue analysis in archaeological pottery samples
Creator:
Matuszewicz, Todd
Place of Publication:
Denver, Colo.
Publisher:
Metropolitan State University of Denver
Publication Date:
Language:
English

Record Information

Source Institution:
Metropolitan State University of Denver
Holding Location:
Auraria Library
Rights Management:
Copyright [name of copyright holder or Creator or Publisher as appropriate]. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
The Methodological Development of Lipid Residue Analysis in Archaeological Pottery
Samples
by Todd Matuszewicz
An undergraduate thesis submitted in partial completion of the Metropolitan State University of Denver Honors Program
May 5, 2017
Dr. April Hill
Dr. Joshua Martin Dr. Megan Hughes-Zarzo
Primary Advisor
Second Reader
Honors Program Director


SENIOR HONORS THESIS The Methodological Development of Lipid Residue Analysis in Archaeological Pottery Samples
T.A. Matuszewicz, D.A. Regan, E. Ball*, and A. A. Hill*
Todd Matuszewicz Metropolitan State University of Denver, De Regan University of Colorado at Denver. Eric Ball and April Hill *Department of Chemistry, Metropolitan State University of Denver, 3058 Science Building P.O. Box 173362. Campus Box 52. Denver, CO 80217-3362 (USA). Room 3057.
ABSTRACT
This research establishes a methodology for extracting fatty acid residue from archaeological pottery sherds in order to verify the use of the vessels and corroborate theories of seasonal migration through dietary habits. Pottery sherds from unidentified sections of the vessels were powdered and washed in solvent to withdraw the embedded fatty residue. Fatty acids are present as free fatty acids or tri, di or monoacylglycerols in all natural animal or plant food sources. (Table 1).
Table 1: Fatty acid composition (%) of free fatty acids, mono,di and triacylglycerols of the tropical plant Mabea fistiilifera.1
Fatty Acids Free Fatty Acids Mfl: Mf2: 38-77 26-90 Monocylglyccrols or Diacylglyccrols MU: Mfl: Mfl: 21-55 28-37 1-27 Triacylglycerols Mf2: Mf3: 1-25 1-20
Cl 4:0 2.6 0.9 0.4 9.5 1.2 1.5 5.8
Cl 6:0 72.2 55.7 31.3 34.1 38.1 14.2 31.9
Cl 8:0 0.4 17.8 0.4 16.5 3.7 1.9 12.2
Cl 8:1 17.2 22.0 8.0 20.2 21.9 20.4 29.7
CI8:2 0.02 1.7 28.0 10.6 33.9 56.0 18.5
Cl 8:3 7.6 1.3 30.0 2.2 0.7 1.0 1.2
C20:0 - 0.5 2.0 6.8 0.4 5.1 0.7
Through various procedures, these fats were mechanically and chemically prepared for analysis. Using computer software, the residues were broadly identified through mass spectra fractional analyses and reference to a spectra library. Further confirmation was achieved through a comparison to lipid ratio decomposition biplot data and archaeological speculation.
The residues were tentatively identified as having originated from a small mammal and
a mix of seeds, nuts and berries. To more precisely identify the source substance, further investigation into modern source materials and techniques for isolating proteins, biomarkers, and isotopes will need to be studied and developed.
INTRODUCTION
There are several approaches accessible to investigate the residue in archaeological pottery and determine source material. The choice of lipids as an analyte was based partly on the limited financial and analytical instrumentation resources of the project, but also on the relative stability of lipid compounds, the ease of analysis; and the various existing methodologies.2
While organic residue analysis is somewhat novel in the field of archaeology, it is clear that the interdisciplinary efforts of archaeologists and chemists will yield an increasingly wide array of results.3
One obstacle that is consistent in this field of study is the preponderance of papers concentrating on the examination of results, while methodologies are brief and imprecise. There is scant effort to collaborate between investigators to crosscheck methodologies or verify the efficacy of procedures.4
The decomposition of source material and the ability to reverse engineer fatty acids from residues is one of the vexing problems of archaeo-chemical analysis. Various initial


Matuszewicz, Regan, Ball and Hill
materials can be manipulated by human as well as natural forces; see Figure 1.
Plant origin
Animal origin
Bitumen* Blood
Amber* Fats
Gums Beeswax
Resins Organic Honey
Tars Residues Dairy
Dyes products
Oils Bread Fermented beverages Proteinaceous glues
Figure 1. Schematic of the origin of the organic remains of natural products potentially preserved in archaeological artifacts: they may be from plant or animal origin and may have been used either as raw materials (above) or after various transformations such
as heating or fermentation (below). The symbol...
indicates a fossil origin.5
The process of simulating oxidation of fatty acids (Figure 2) and experimentally recreating residues in ceramic matrices was also investigated in this paper. The issue of food degradation has relevance in food science as well as in paleontology.6 7 Almost all the work has been done exclusively in the former. However, the timescale for archaeological decay is much larger than those for food science and the factors leading to degradation and rancidification more variable.8"10
This paper will explicitly outline techniques for the destructive pottery analysis of lipid residue in archaeological samples.11'13 Although research has begun on other ancient artifacts such as fire-cracked rock or abdominal content from gravesite-specific soil samples, this paper is limited to ancient ceramics.1415 Various methods of extraction, esterification, derivatization and silylation were explored,16"18 this analysis is from these studies and could be an initial attempt
to assemble a methodological database for the standardization of sample preparation practices modeled after spectral databases such as the Spectral Database for Organic Compound (SDBS).
9
CH.0-0
I 9
1. 2-diacylglycerol 1, 3-diacylglycerol
-RHdo,H
* o iS
.1 CH,OH | o
i CHOH 1 1 i > i CHO-C--~ |
i CH,OH J CH.OH
1 -monoacylglycerol 2-monoacylglycerol
Hydrolysis products
CH,OH ?
CHOH HO-C'-~----
CH,OH Free fattV acjcl
Glycerol
Figure 1. Hydrolytic pathway for the transformation of triacylglycerols To free fatty acids.19
MATERIALS and METHODS Pottery Samples
Samples were obtained from four different sources. For the surrogate preparations commercial pottery was initially used,10 however, subsequent surrogates replicated traditional clay matrixes and manufacturing techniques. Field sample sherds, obtained from Dr. D. Hill, and authentic reproductions from the Swink Pottery Studio were employed. Half of the samples were used raw and the other half were ashed at 450 C for 17 hours. As shown in Figure 3, the ashing changed the inner composition of the samples.
2


Matuszewicz, Regan, Ball and Hill
Figure 3. A) A raw Swink archaeo-pottery sherd. B) An ashed Swink archaeo-pottery sherd heated at 450 C for 17 hours.
Of the four archaeological samples, only one had any identifying marks (FS#1082.1, Bg#1082, Proj.2033, LA182820, 10-28-2016).
Oil Samples
Four different oils were used as initial food source material. These consisted of avocado, olive, coconut oils, and lard. All the oils were obtained commercially and used as received. The samples were stored at room temperature. The coconut and lard were melted at 60 C before applying and sampling.
Sample Preparation
Sherds, prepared as either ashed or raw, were coated with the various oils. The first set, or group, of sherds contained no additional oil and was used as a blank. The other sherds were saturated in oil. Due to the difference in porosity between the samples, the amount of oil applied varied. All of the samples were coated until an even absorption of oil was observed. The second set remained unheated. The third set of sherds were heated in an oven at 250 C for 1 hour. The fourth set of sherds heated 1 hour at 250 C, cooled 1 hour to room temperature, and heated 1 hour at 250 C. The fifth set of sherds were heated lhour at 250 C, cooled lhour to room temperature, and cooked overnight at 250 C. The sixth
set of sherds cooked overnight at 75 C (Table 2).
Table 2: The preparation of various sample sets for degradation simulations.
Sample Coated with oil Cooked lhour at 250C Cooled to Room temp. Cooked lhr at 250C Cooked overnight at 250C Cooked overnight at 75C
Set 1 No No NA NA NA NO
Set 2 Ye s No NA NA NA No
Set 3 Yes Yes Yes No No No
Set 4 Yes Yes Yes Yes No No
Set 5 Yes Yes Yes No Yes No
Set 6 Yes No NA Yes No Yes
Chemicals
Solvents used in the extraction and preparation of the oils from the prepared sherds included dichloromethane, chloroform, methanol, hydrochloric acid, petroleum ether and iso-octane ( 2,2,4-trim ethylpentane). All solvents were purchased from Fischer Scientific (Waltham, MA, USA) except the petroleum ether which was purchased from Mallinckrodt (Staines-Upon-Thames, Surrey UK). The 0.5 N anhydrous hydrochloric acid and N,0-bis(trimethylsilyl) trifluoro-acetamide (BSTFA) + Trimethyl -chlorosilane (TMCS), 99:1, a column evaluation FAME mix and a C8-C24 FAME standard mix were purchased from Supelco (Bellefonte, PA, USA).
Extraction
Lipids were extracted using a combination of methodologies.20"22 Samples = 1 cm2 were crushed with a mortar and pestle. A 400 mg portion of each powdered sample was placed in a test tube and mixed with 5 mL of a 2:1, (v:v.) mixture of chloroform and methanol or dichloromethane and methanol. Samples were sonicated for 10 minutes and centrifuged for 10 minutes. The supernatant
3


Matuszewicz, Regan, Ball and Hill
was syringe filtered through a Thermo Scientific 0.20 pm PDVF membrane filter and transferred to a vacuum centrifuge tube. The extraction process was repeated with a 15 second vortexing of the sample/solvent solution before sonication and centrifuging. The 10 mL aliquot of lipid/solvent was vacuum centrifuged 120 minutes at 37 C until less than 1.0 mL of analyte remained. Samples were transferred to a screw-top glass vial with a Teflon septum cap and flushed with nitrogen (N2) before storage in a -20 C freezer.
Free Fatty Acid Derivatization
Several 400 pL aliquots of the total lipid extract/solvent were vacuum centrifuged until dry. The dried lipids were transferred into Teflon septum screwtop glass vials. Fatty acid methyl esters (FAMEs) were prepared and derivatized through two methods; see Figure 4.The first by adding 3.0 mL of 0.5 N anhydrous hydrochloric acid in methanol and heating for 60 min at 70 C. The second used 4.1 ml of 0.5 M methanolic HC1 and was heated for 60 minutes at 68 C in a heating block. These methods ensured that any fatty acids in the sample still attached to the glycerol backbone as tri-, di- or monacylglycerols (TAGs, DAGs, or MAGs) were detached and converted to FAMEs (esterification). After cooling to room temperature, the FAMEs were transferred to a test tube with 3x1 mL of 18 MQ water. The FAMEs were retrieved with 2.5 mL of petroleum ether. A 1000 pL Finn Pipette II was used to remove the upper solvent layer and transfer it to a crimp-top GC vial. The solution was dried in a vacuum centrifuge (=15 min.) at 37 C to remove the solvent. The dried FAMEs were dissolved with 75 pL of iso-octane and
transferred to a GC vial with a 100 pL conical insert.
Figure 4. The molecular structure of a fatty acid methyl ester FAME and the mechanism for esterification.23
Silylation Preparation of BSTFA and TMCS Derivatives
A 100 pL aliquot of the total lipid extract solution was transferred to a Teflon septum screw-top 4 mLvial. Trimethylsilyl (TMS) derivatives were prepared by adding 70 pL N,0-BSFTA and TMCS 99:1, (v:v), and heating in a heating block for 30 minutes at 70 C. After cooling to room temperature,
75 pL of the derivative was transferred to a GC vial with a 100 pL conical insert.
GC/MS Analysis
The GC analysis was performed on a Hewlett Packard (HP) 6890 gas chromatograph fitted with a Hewlett Packard 5973 mass selective detector, a 7683 series auto injector, and connected to a personal computer equipped with ChemStation software. Samples were run through a Zebron ZB5-HT column with a (5%-Phenyl)-methylpolysiloxane phase
4


Matuszewicz, Regan, Ball and Hill
specifically designed for the high-quality separation of fatty acid methyl esters (FAMEs). The ZB5-H is a silica fused capillary ( 30 m x 0.32 mm I.D. 25 pm; Phenomenex, Torrance, CA,USA). An autosampler injected a 3 pL volume using a splitless injection system. Helium was the carrier gas with a column flow rate of 1.0 mL/min. Two different GC programs were run. The first method involved an inlet temperature of 250 C and the initial column temperature was 100 C; this was increased at a rate of 12 C per minute to a final temperature of 325 C, which was held for 12 minutes. The second method involved the same column and instrument but had a slower ramp. The initial temperature was 140 C rising to 250 C at 6 C/minute with a 5 minute hold at the end. The chromatogram peaks were integrated using HP ChemStation software.The mass spectra were compared against the National Institute of Standards and Technology (NIST) Mass Spectral library (1998). Solvents, chemicals, pottery stock, and the column were checked for impurities by running blank samples before, during, and after each group of samples.
Results
Derivatizations of Neat Oils
The first attempt at derivatization was the use of the neat oils avocado, olive, coconut and lard. Random quantities of the fats, between 100 and 800 mg, were heated at 60 C for 4 hours 45 minutes with 3 mL of methanolic HC1 in a Teflon septum screw top vial. The results of the FAME abundances correlates to the lipid ratios of the raw oils, see Appendix A. The lipids follow the nomenclature system of C 16:1
where C is carbon, the first number is numbers of carbons in the molecule, colon, and the second number is number of double bonds or degrees of unsaturation, see Table 3 for ratios. The extraction and derivatization techniques were not subject to analysis as the neat samples provided abundances too large to assess extraction and derivatization.
NEAT OILS Fast Ramp GC Peak Area Percentage
Sample C12:0 C14:0 06:1 06:0 08:2 08:1 08:0 C20:l C20:0
Avocado Oil
Percentage 12.50% 21.68% 65.82%
Literature Values 2-10% 12-20% 9-17% 55-75%

Coconut Oil
Percentage 34.84% 22.57% 17.41% 18.25% 6.94%
Literature Values 44-52% 13-19% 8-12% 34-62% 19-49% 2-5%

Olive Oil
Percentage 5.55% 23.96% 64.89% 4.72% 0.88%
Literature Values 7.5-20% 3.5-21% 55-83% 0.5-5%

Lard
Percentage 0.48% 0.12% 6.74% 27.31% 0.95% 43.93% 14.70% 2.10% 0.41%
Literature Values 1% 3% 25-28% 6-10% 44-47% 12-14%

Surrogates were prepared using pottery sherds obtained from three sources: Home Depot,Inc., Swink Pottery Studio, and field samples from Costa Rica obtained from Dr. D. Hill which registered no lipid residue on previous analysis. Half of the samples were ashed in an oven at 450 C for 24 hours to destroy any organic residues. Lipids begin to rapidly denigrate at 80 C and with higher temperatures any incidental contamination is removed. This process creates a clean sherd that is the platform for oils to be artificially degradated in an attempt to model
Table 3. The GC/MS analysis of neat oils using fast temperature ramping in the column parameters.
The four oils: avocado, coconut, olive and lard, were derivatized and analyzed and the results were compared against literature data.
Surrogates and Degradations
5


Matuszewicz, Regan, Ball and Hill
environmental degradation. The other half were left unheated so as not to modify the traditional unfired carbon core of archaeological pottery. The four neat oils were spread on three of the ashed sherds, one from each of the suppliers. The first group of three sherds were left as is with raw oil. The second group of three sherds were heated at 250 C for 1 hour. The third group of three sherds were heated 1 hour at 250 C, allowed to cool to room temperature for 4 hours and then heated another hour at 250 C. Although each of the oils provided results, olive oil became the preferred oil of investigation for practical reasons of simplification. The results of this inquiry are displayed in Table 4.
Table 4. Olive oil on two ashed sherds from a previous investigation in Costa Rica. The uncooked (raw) samples exhibit traditional lipid ratios for olive oil. The cooked samples have no lipid residue remaining in them. Although four oils were investigated, only the olive was utilized as a standard.
Sample Fast Ramp GC
COSTA RICA 06:0 08:1 08:0
Raw7 ( Peak Area) 139853677 338776981 172707230
21.48% 52.00% 26.52%
2 Hrs @ 250C 0 0 0
The chromatograms are of olive oil uncooked and heated for 2 hours at 250 C are exhibited in Figure 5 and Appendix A. A second degradation method was employed on the archaeo-pottery sherds from Swink Pottery. As expected, the ashing changed the inner matrix of the sherds; see Figure 3. Sherds were analyzed without lard, with lard but uncooked and with lard and cooked at 75 C for 48 hours; see Table 5.
Figure 5. The chromatogram for the fast ramp GC program of methyl ester derivatized olive oil uncooked on an ashed surrogate pottery sample.
Table 5. Total Lipid Extraction (TLE) of various fatty acids in surrogate pottery samples. The top group is the raw sherds with no lard, with lard but uncooked, and with lard and cooked at 75C for 48 hours. The loss of abundance is shown as a percentage loss. The lower group is the ashed sherds with no lard, with lard but uncooked and with lard and cooked at 75C for 48 hours. The loss of abundance is shown as a percentage loss. The difference in degradation is shown as a ratio. The raw degrades faster and mono-unsaturated oils (MUFAs) degrade faster than saturated oils (SFAs).
Sample Fast Ramp GC ( Peak Area)
Swink Pottery 04:0 06:1 06:0 08:1 08:0
Raw Sherd 0 0 6132503 5462884 6086934
Sherd with Lard 64079117 77848227 695047570 922806480 368025200
Difference 64079117 77848227 688915067 917343596 361938266
48 Hrs @ 75C 27582806 0 288900065 208455705 130914551
Difference 36496311 77848227 400015002 708887891 231023715
% Change 56.96 100.00 58.06 77.28 63.83

Ashed Sherd 0 0 7062001 2082511 4177965
Sherd with Lard 38036811 404513226 435942173 455908097 202886050
Difference 38036811 404513226 428880172 453825586 198708085
48 Hrs @ 75C 26086887 0 269157846 128819205 126286549
Difference 11949924 404513226 159722326 325006381 72421536
% Change 31.42 100.00 37.24 71.61 36.45

Degradation
Raw vs Ashed 1.81 1.00 1.56 1.08 1.75

SFAs vs MUFAs 1.70:1.00
6


Matuszewicz, Regan, Ball and Hill
Extraction and Derivatization Method
Several different extraction and derivatization techniques were investigated over the course of the project. The multiplicity of variations is beyond the scope of this paper. In brief, it was concluded that a chloroform methanol solvent 2:1 (by volume) was the most practical and effective at extracting lipids.
The use of methanolic HC1 to derivatize the fatty acids into FAMEs provided the clearest and most useful data. None of the other methodologies were effective and the unusable data was left out of this analysis.
The final extraction and derivation procedure is: The Extraction of Lipid Residue
1. Crush sample with mortar and pestle and transfer 400 mg to a test tube.
2. Powdered sample is mixed with a 2:1 (by volume)mixture of chloroform and methanol (5 ml).
3. Ultrsonicated (10 min).
4. C entrifuged (10 min).
5. Filter solids from solvents through syringe filter into a vacuum centrifuge test tube.
6. Repeat steps 2-5.
7. Vacuum centrifuge at 37 C until less than 1.5mL of analyte remains.
(Sample can be transferred to a Teflon screw top vial, flushed with nitrogen and stored in -20 C freezer).
Preparation of Fatty Acid Methyl Esters FAMES
8. A 400 pL aliquot of lipid solution is placed in a GC vial in a vacuum centrifuge tube. Vacuum centrifuge at 37 C until dry. Approximately 15 min.
9. Fames are prepared by treating the dry lipid with 4.0 mL of 0.5 M methanolic HC1.
10. Heat in heating block in capped Teflon screw top vials for 60 min at 68 C.
11. Vortex for 15 sec.
12. After cooling to room temp, transfer to test tube using 3.4 mL of ultrapure water.
13. FAMES are recovered using petroleum ether (2.5 mL).
14. Transfer the top layer to a GC vial using a 1000 pL autopipette.
15. Solvent is removed using heat. Vacuum centrifuge at 37 C until dry)
16. FAMES are dissolved with 75 pL of isooctane.
17. Transfer to GC vial with conical insert.
An illustrated detailed procedure is outlined in Appendix B.
GC/MS Programs
Two separate GC/MS programs were used.
The fast temperature ramp was the first method and the slow ramp was the second.
The retention times for various FAMEs are listed in Tables 6 and 7 respectively. The slow ramp chromatogram of fatty acid methyl ester standards is shown in Figure 6.
Table 6. Retention times of fatty acid methyl esters (FAMEs) using a fast temperature ramp GC/MS program.
C12:0 C13:0 C14:0 C15:0 C16:l C16:0
6.629 8.506 8.998 9.376 10.087 10.337

C17:0 C18:2 C18:l C18:0 C20:l C20:0
10.815 11.741 11.786 11.976 13.145 13.311
Table 7. Retention times of fatty acid methyl esters (FAMEs) using a slow temperature ramp GC/MS program.
C8:0 00:0 02:0 04:0 06:1 06:0
3.445 6.14 9.622 13.202 16.224 16.591

08:2 08:1 08:0 C20:0 C22:l C22:0 C24:0
19.211 19.327 19.712 22.617 25.01 25.365 28.792

7


Matuszewicz, Regan, Ball and Hill
Figure 6. The chromatogram for the slow ramp GC program of fatty acid methyl esters.
Archaeological Samples
Four archaeological samples and the dirt matrix from which they were excavated were analyzed. Two of the archaeological samples showed no significant difference between the total lipids of the soil and the sherds. Those two samples are not included in this analysis. The total lipid extracts of two other specimens, pot sherds #1 and #4 are in Table 8. Pot sherd #1 did not have sufficient residue to warrant analysis.
Table 8. Total Lipid Extraction (TLE) from two archaeological samples, pot sherd #1 and pot sherd #4. The analysis used a slow ramp GC/Ms program. Abundances were measured in peak area. Sherds #2 and #3 had negligible TLE.
Sample C14:0 C16:0 06:1 08:0 08:1
#1 Pot 550597 2982624 633174 1515463 791539
#1 Dirt 500375 2823786 1340313 1457141 2656892
Difference 50222 158838 0 58322 0

#4 Pot 701164 4084422 1785337 1714343 1917337
#4 Dirt 684684 3434403 1390562 1568203 1489058
Difference 16480 650019 394775 146110 428279
Discussion
Extraction and Derivatizations
Methods other than the one from the methods section, which is further detailed in Appendix B, did not produce any useful
results. The chromatogram in Appendix A shows the results of one of the silylation derivatizations. Given the considerable time involved in adapting the esterification derivatization, a silylation methodology is beyond the prevue of future research in this lab. The extraction method simplified previous methods, eliminating the need for expensive chemicals (0.5 N anhydrous hydrochloric acid) in step 9 and the use of a nitrogen glove box in steps 7, 8 and 15. The method did leave questions about the efficacy in regards to the identification of medium chain FAMEs and very long chain FAMEs. It is not clear whether this methodology can effectively extract and derivatize lipids medium and very long chain compounds.
Surrogates and Degradations
The Swink archaeo-pottery sherds were not compared against the Home Depot sherds or previously negated field samples. The Swink sherds were analyzed using a different degradation method than the other two. The second group of samples returned no viable data due to overly high temperatures in the degradation process. Without a formal analysis, it is clear from Table 4 and Appendix A that the degradation procedure of 250 C destroyed any lipids. In the archaeo-sherds, there was considerable difference between the degradation of fatty acids in the raw sherds versus the 'ashed' sherds (Table 5). Lipids 04:0, 06:0, and 08:0 degraded 1.81, 1.56, and 1.75 times faster, respectively, in the raw sherd than in the ashed sherd. Unquestionably, the composition of the surrogate sherds affects the decomposition rate of lipids. The mono-unsaturated fatty acids were more unstable than the saturated lipids (Table 5). The greater the number of double bonds in a fatty acid, the greater the oxidation
8


Matuszewicz, Regan, Ball and Hill
susceptibility due to the addition of more methylene- interrupted carbon reaction sites. This trend is in agreement with previous studies (Figure 7). 06:1 and polyunsaturated fatty acids were absent in the degraded samples. This suggests a necessary adjustment to future methods.
Figure 7. Simulated degradation for four fatty acid ratios over 100 decomposition steps.24
GC/MS Programs
By comparing the two separate GC/MS temperature programs, the slow ramp allows for the identification of the medium chain fatty acids, C8:0 02:0, and the very long chain C22:0 C24:0. Although the fast ramp method seems to indicate an advantage in the identification of odd long chains, it is difficult to compare due to the fact that those lipids were not present in the standards used for the slow ramp analysis (Tables 6 and 7). The slow ramp is preferable as it allows for clearer results through better resolution of the peaks.
Archaeological Samples
Only one of the archaeological samples provided sufficient lipid residue for analysis, number 4. By using a lipid ratio analysis (Table 9) and comparing it against the lipid biplots from literature in Figures 8 and 9, the composition appears to have originated from terrestrial mammal or seeds. These biplots
were generated from previous research in the western regions of Canada.21By drawing a vertical line for the numerical ratio of the lipids C16:l/C18:l and a horizontal line for the ratio of (C15:0 + C17:0)/C 18:0, the intersection designates the origin source. In Figure 8 that would be roots or seeds. Figure 9 is another example of a biplot ratio graph using different lipids.
Table 9. Analysis of long chain fatty acids (LCFAs) in archaeological pottery sherd #4.
LCFAs Percentage of Lipitl Residue Lipid Ratios Ratios Amounts
04:0 13.4% (C12:0 + C14:0) (C20:0 + C22:0)
06:0 11.3% tC16:0 + C16:1) (C18:0 + C18:1) 0.181
06:1 2.0% C16:1 C18:1 0.033
08:0 13.7% C16:0 08:0 0.823
08:1 59.7%
Product:
seed
* root meat c green fish
p berry
'o
C16:1/C18:1
Figure 8. Biplot of two conservative fatty acid ratios for modern food products (ellipses represent subjective estimations of the range of values for each food group). The vertical line represents the known lipid ratio of 06:1:08:1.17
9


Matuszewicz, Regan, Ball and Hill
Product:
* seed root meet
* green
* fish berry
0 2 4 10 12 M It
C16.1 / 018:1
Figure 9. Biplot of two additional conservative fatty acid ratios for modern food products (ellipses represent subjective estimations of the range of values for each food group). The vertical line represents the known lipid ratio of 06:1:08:1.17
Table 11. Analysis of individual lipid percentages of the Total Lipid Extract (TLE) from pottery sherd number 4. These percentages confirm the presence of mammal and a mix of seeds, nuts and berries.10
Cluster A B C
Subcluster I n in IV V VI VII vm K X XI XD xm XIV XV
Tjpe Manunal Fat and Marrow Large Herbivore Mai Fish Fish Bnies and Nuts Mixed Seeds and Bernes Roots Seeds Mind Greens Berries Roots Greens Roots
06:0 19.90 19.1) 16.01 14.10 3.13 0 1.1! 19.9! 7.52 10.33 18.71 3.17 22.68 21.19 18.71
08:0 20.35 1.8? 11! 1.47 2.36 138 2.59 3.33 2.13 2.48 1.34 3.13 3.66 3.91
08:1 35.79 1828 31.96 31.11 33.29 29.1! 6,35 10.0! 13.6! 3.03 14.95 12.1! 4.05 3.34
08:2 7.01 1,91 191 1.01 11.11 33.83 31.69 18.14 61.11 39.21 18.8! 29.08 26.21 16.15 15.61
08:3 0.68 161 139 3.83 1.03 3.(6 1.31 7.24 3.49 19.7? 35.(8 39.75 9.61 17.81 3.4!
YLCS 0.16 0.32 023 0.13 0.16 4.46 2.98 8.50 5.19 3.13 6.77 9.10 15.3! 18.(8 43.36
via 0.77 4.29 39.91 24.11 0.23 2.10 1.00 2.23 0.99 2.63 1.13 0.95 2.06 0.7! 1.10
V1CS- Veil' Long Chain (C20, C2! mi CHI Sound Fitly Adds MCU Voy Long Chain (C20. C22 and C24) liisaniated Fatty Acids
Further lipid ratio analysis confirms the same two food groups sources (Table 10).
Table 10. Distinguishing food types by fatty acid ratios (after Eerkens 2001,106). Some seeds and nuts have values well over 4, making this a worthwhile ratio. The evidence from sherd #4 indicates that the source material was likely mammal and seeds and nuts.
Ratios Mammal meat and fat Fish Bird Roots Greens Seeds and nuts Berries
]C12:Q+C14:0) (C20:0*C22d) >3.5 >5 >3 <2 <2 <2 <0.8
rC16:0+C16:1) (C18:0+C18:1) I I <1.5 >0.5 and <0.8 >1.5 >2 <0.5
SIM C18.1 >0.2 and <0.7 1 41 1 <1 >1 <1.1 or >0.1 and <03 <0.2
C16:0 C18:0 I <* I >3 <3 >3 >3 : ? e' H
The last analysis of the lipids in sherd number 4 is a straight percentage of various lipids detected in the analysis. The results indicate the lipid sources to be from terrestrial mammal and a mix of seeds, nuts and berries (Table 11).
Further Research
This project marks the initiation of exploration in to the field of archaeo-chemistry. Future experiments can provide more information on the extraction and derivitization of medium and very long chain fatty acids. More methods for the degradation of lipids, including light and moisture, and new techniques of heating samples should beinvestigated. A database of regional specific food sources and their lipid composition after preparation is needed to more precisely identify origin materials. The investigation of biomarkers, isotopic relationships, yeast residues, proteins and electon microscopy are all potential additions to the lipid research.25"28
10


Matuszewicz, Regan, Ball and Hill
CONCLUSION
With a simplified extraction/derivatization technique, the use of archaeo-pottery samples as surrogates, a database of regional source materials, the lipid composition of those source materials and an artificial degradation model, the possibility of historical insights into dietary, medicinal and cultural practices through chemical analysis is unlimited.1518 27 This paper has outlined a simple and specific methodology
that uses inexpensive chemicals and equipment and can be easily learned by a novice chemist. The GC temperature program outlined in this work produces clear peak resolution. With these standardized techniques the collection of greater source origin material lipid compositions is possible. Combined with continued investigation into lipid degradation, the analyses of archaeochemistry will become more accurate.
Acknowledgements
We would like to thank David Hill for the opportunity to work on the archaeological food residue project, the Undergraduate Research Department of Metropolitan State University of Denver provided funding for the GC/MS analysis and Clint Swink for providing traditional Anasazi pottery for surrogate preparation.
References
1. Scientific Electronic Online Library http://www.scielo.br/scielo.php7scrip t=sci_arttext&pid=SO 100-40422008000300002 (accessed May 11, 2017).
2. Murphy, R.C. Mass Spectrometry of Lipids. Handbook of Lipid Research Volume 7. 1993 New York: Plenum Press.
3. Eerkens, J.W. and Barnard, H. Introduction. The Theory and Practice of Archaeological Residue Analysis. Introduction. 2007 Oxford, England: Archaeopress.
4. Barnard, H.; Dooley, A.N. and Fauli, K.F. An Introduction to
Archaeological Lipid Analysis by Combined Gas Chromatography Mass Spectrometry (GC/MS). The Theory and Practice of Archaeological Residue Analysis. Chapter Five. 2007 Oxford, England: Archaeopress.
5. Regert, M.; Gamier, N.; Decavallas, O.; Cren-Oliv, C. C. and Rolando, C Structural characterization of lipid constituents from natural substances preserved in archaeological environments. Meas. Sci. Technol. 2003, 14, 1620-1630.
6. Collins, M.J.; Nielsen-Marsh, C.M.; Hiller,V; Smith, C.I.; Roberts, J.P.;
11


Matuszewicz, Regan, Ball and Hill
Prigodich, R.V.; Wess, T.J.; Csapo,
J.; Millard, A.R. and Turner-Walker,
G. The Survival of Organic Matter in Bone. A Review. Archaeometry 2002 44,3: 383-394.
7. Frankel, E.N. Lipid Oxidation. 1998 Ayr: Oily Press.
8. Evershed, R.P. Biomolecular Archaeology and Lipids. World Archaeology 1993 25(l):74-93.
9. Malainey, M. E. Lipid Residue Analysis. Manuals in Archaeological Method, Theory and Technique. Chapter 14. 2011, pp.201-217. New York,
Dordrecht, Heidelberg, London:
Springer Press.
10. Malainey, M. E.; Przybylski, R. and Sherriff, B. L. The Fatty Acid Composition of Native Food Plants and Animals of Western Canada. J.
Archaeol. Sci. 1999, 26, 83-94.
11. Evershed, R. P. Biomolecular Analysis by Organic Mass Spectrometry. In Modern Analytical Methods in Art and Archaeology,
2000 edited by E. Ciliberto and G.
Spoto, Volume 155: 177-239.
Chemical Analysis. John Wiley &
Sons, New York.
12. Craig, O. E.; Steele, V. J.; Fischer, A.;
Hartz, S.; Andersen, S. H.; Donohoe, P.; Glykou, A.; Saul, H.; Jones, D. M.;
Koch, E. and Heron; C. P. Ancient lipids reveal continuity in culinary practices across the transition to agriculture in Northern Europe. Proc. Natl. Acad. Sci.
U. S. A.2011, 108, 17910-17915.
13. Shishlina, N.I. The Seasonal Cycle of Grassland use in the Caspian Sea Steppe during the Bronze Age. A New Approach to an Old Problem. European Journal of Archaeology
2001 4,3: 346-366.
14. Berg, G.E. Last meals. Recovering Abdominal Content from Skeletonized Remains. Journal of Archaeological Science 2002 29: 1349-1365.
15. Quigg, M.J.; Malainey, M.E.; Przybylski, R. and Monks, G. No Bones about it. Using Lipid Analysis of Burned Rock and Groundstone Residue to examine Late Archaic Subsistence Practices in South Texas. Plains Anthropologist 2001 46. 283-303.
16. Aveldano, M.I., VanRollins, M. and Horrocks, L. A. Seperation and Quantation of Free Fatty Acids and Fatty Acid Methyl Esters by Reverse Phase High Pressure Liquid Chromatography. Journal of Lipid Research 1983 24:83-93
17. Eerkens, J.W. Organic Residue Analysis and the Decomposition of Fatty Acids in Ancient Potsherds.
The Theory and Practice of Archaeological Residue Analysis. Chapter Eight. 2007, Introduction. Oxford, England: Archaeopress.
18. Patrick, M.; de Konig, A. J. and Smith, A. B. Gas Liquid Chromatographic Analysis of Fatty Acids in Food Residues from Ceramics Found in the Southwestern Cape, South Africa. Archaeometry 1985, 27(2): 231-236.
19. Frankel, E. N. Recent Advances in Lipid Oxidation. J. Sci. Food Agric. 1991, 54, 495-511.
20. Eerkens,J.W., British Archaeological Reports International Series 1650. pp.42-60. Oxford, UK.
21. Folch, J.; Lees, M.; and Sloane Stanley, G. H. A simple method for the isolation and purification of lipid extracts from
12


Matuszewicz, Regan, Ball and Hill
brain tissue. J. ofBio.l Chem. 1957, 191:833.
22. Malainey, M. E.; Malisza, K.L.; Przybylski, R. and Monks, G. The Key to Identifying Archaeological Fatty Acid Residues. Paper presented at the 34th Annual Meeting of the Canadian Archaeological Association, Banff, Alberta, May 2001
23. Evershed, R. P.; Dudd,M.; Copley, M.S.; Berstan, R.; Stott, A.W.; Mottram, H.; Buckley, S.A. and Crossman, Z. Chemistry of Archaeological Animal Fats. Accounts of Chemical Research 2002 35:660-668.
24. DeMan, J.M. Chemical and Physical Properties of Fatty Acids. In C.K. Chow (ed.) Fatty acids in foods and their health implications. 2008 CRC PressBoca Raton., pp. 17-45.
25. Barnard, H. and Eerkens, J.W. Abstracts for the Symposium on Theory and Practice of Archaeological Residue Analysis at the 70th Annual Meeting of the
Society for American Archaeology, Salt Lake City (Utah), March 2005.
26. Budja, M. Neolithic pottery and the biomolecular archaeology of lipids. Documenta Praehistorica 2014 41:196.
27. Evershed, R. P.; Heron, C. and Goad, L. J. Analysis of Organic Residues of Archaeological Origin by High Temperature Gas Chromatography and Gas Chromatography-Mass Spectroscopy. Analyst 1990 115:1339-1342.
28. Wang, J.; Liu, L.; Ball, T.; Yu, L.;
Li, Y. and Xing, F. Revealing a 5,000-y-old beer recipe in China. Proceedings of the National Academy of Sciences 2016 113, 6444-6448.
13


Matuszewicz, Regan, Ball and Hill
Appendix A:
File : C:\HPCHEM\1\DATA\ARSON\AC2.D Operator : April Hill
Acquired : 12 Oct 2016 11:53 using AcqMethod ARCCHEM Instrument: GC/MS #1 Sample Name: Avocado Oil Misc Info :
Vial Number: 2
56+07
4.5e+07
4e+07
3.5e+07
TIC: AC2.D
36+07
2.56+07
2e+07
1.56+07
16+07
, A| I | i*i i I Wfin I
6.00 7.00 8 00 9 00 10.00 11.00 12.00 13.00
jifj |i(..-|i.r.| .y. i .. ,. i. |, i r | ., 11 j .'7T7~TT7 r"T prrrr;--rr:. ,---77-7-.-4 00 1500 16 00 17!00 18.00 1900 20.00 21.00 22.00 230 0 24.00 2500 2600 27 00 28.00 29 00 30 00
Figure 10. The chromatogram for the fast ramp GC program of methyl ester derivatized neat avocado oil
File : C:\HPCHEM\1\DATA\ARSON\AC1.D Operator : April Hill
Acquired : 12 Oct 2016 11:18 using AcqMethod ARCCHEM Instrument: GC/MS #1 Sample Name: Coconut Oil Misc Info :
Vial Number: 1
5e+07
4.56+07
4e+07
3.56+07
3e+07
2.5e+07
26+07
1.56+07
16+07
5000000
TIC: AC1.D
l-|l| ir
6.00 7.00

8.00 9.00 10.00 11.00 12.00 13.00
14.00 15.00 16.00 17.00 18 00 19.00 20.00 21.00 22.00 23 00 24.00 25 00 26.00 27.00 28.00 29.00 30.00
Figure 11. The chromatogram for the fast ramp GC program of methyl ester derivatized neat coconut oil
14


Matuszewicz, Regan, Ball and Hill
File : C:\HPCHEM\1\DATA\ARSON\AC3.D Operator : April Hill
Acquired : 12 Oct 2016 12:29 using AcqMethod ARCCHEM Instrument: GC/MS #1 Sample Name: Olive Oil Misc Info :
Vial Number: 3
Figure 12. The chromatogram for the fast ramp GC program of methyl ester derivatized neat olive oil
File : C:\HPCHEM\1\DATA\ARSON\AC4.D Operator : April Hill
Acquired : 12 Oct 2016 13:04 using AcqMethod ARCCHEM Instrument: GC/MS #1 Sample Name: Lard Misc Info :
Vial Number: 4
Figure 13. The chromatogram for the fast ramp GC program of methyl ester derivatized neat lard.
15


Matuszewicz, Regan, Ball and Hill
File : C:\HPCHEM\1\DATA\ARS0N\TM13.D Operator : Todd Matuszewicz
Acquired : 3 Feb 2017 16:13 using AcqMethod ARCCHEM Instrument: GC/MS #1 Sample Name: AC 2T1CR2C Misc Info :
Vial Number: 20
Figure 14. The chromatogram for the fast ramp GC program of methyl ester derivatized olive oil cooked on an ashed surrogate pottery sample at 250C for 2 hours..
TIC: TM106.D
1.2e+07
1.1e+07-
1e+07
9000000
6000000 5000000j 400000o] 3000000 2000000 1000000
Mr
4 00 5.00 6.00 7.00 8.00 9.00 10.00 11 00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 19.00 20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00 28.00 29.00
Figure IS. The chromatogram for the slow ramp GC program of fatty acid methyl esters.
16


Matuszewicz, Regan, Ball and Hill
File : C:\HPCHEM\1\DATA\ARS0N\AC13.D Operator : Todd Matuszewicz
Acquired : 26 Jan 2017 16:32 using AcqMethod ARCTMS Instrument: GC/MS #1 Sample Name:
Misc Info :
Vial Number: 23
Figure 16. The chromatogram for the fast ramp GC program of silylation derivatized olive oil uncooked on an ashed surrogate pottery sample.
17


Matuszewicz, Regan, Ball and Hill
Appendix B: Extraction and Derivatization Method
1. Crush sample with mortar and pestle and transfer to a test tube (13 x 100mm).
2. Powdered sample is mixed with a 2:1 mixture v:v of chloroform and methanol (5 ml).
3. Ultrsonicated (10 min).
4. C entrifuged (10 min).
18


Matuszewicz, Regan, Ball and Hill
5. Filter solids from solvents through syringe filter (0.2pm) into a vacuum centrifuge test tube (15 x 85mm lime glass).
6. Repeat steps 2-5.
7. Vacuum centrifuge at 37C until less than 1.5mL of analyte remains 2 hours).
19


Matuszewicz, Regan, Ball and Hill
20


Matuszewicz, Regan, Ball and Hill
Preparation of Fatty Acid Methyl Esters FAMES
8. A 400pL aliquot of lipid solution is placed in a GC vial which is placed in a vacuum centrifuge
tube.
Vacuum centrifuge at 37C until dry = 15minutes.
21


Matuszewicz, Regan, Ball and Hill
22


Matuszewicz, Regan, Ball and Hill
9. Fames are prepared by treating the dry lipid with 4.0 mL of 0.5N anhydrous hydrochloric acid in methanol or 0.5 M HCl/MeOH in a heating block capped Teflon screw top vials (68C; 60min).
10. Vortex 15 seconds.
23


Matuszewicz, Regan, Ball and Hill
11. After cooling to room temp, transfer to test tube. 3.4mL of ultrapure water is added.


Matuszewicz, Regan, Ball and Hill
25


Matuszewicz, Regan, Ball and Hill
12. FAMES are recovered using 2.5mL petroleum ether. Using a 1000pL autopipette pull out 2000pL, the top layer,transfer to a GC vial.
26


Matuszewicz, Regan, Ball and Hill
13. Solvent is removed using heat. Place GC vial in vacuum centrifuge tube. Heat at 37C until dry = 15 minutes.
27


Matuszewicz, Regan, Ball and Hill
28


Matuszewicz, Regan, Ball and Hill
29


Matuszewicz, Regan, Ball and Hill
14. FAMES are dissolved with 75 pL of iso-octane added to dry GC vials using a lOOpL
autopipette.
30


Matuszewicz, Regan, Ball and Hill
15. Transfer to GC vial with conical insert using the same autopipette.
31


Matuszewicz, Regan, Ball and Hill
32


Matuszewicz, Regan, Ball and Hill
33


Full Text

PAGE 1

The Methodological Development of Lipid Residue Analysis in Archaeological Pottery Samples by Todd Matuszewicz An undergraduate thesis submitted in partial completion of the M etropolitan State University of D enver Honors Program May 5, 2017 Dr. April Hill Dr. Joshua Martin Dr. Megan Hughes Zarzo Primary Advisor Second Reader Honors Program Director

PAGE 2

SENIOR HONORS THESIS The Methodological Development of Lipid Residue Analysis in Archaeological Pottery Samples T.A. Matuszewicz D.A. Regan, E. Ball*, and A.A. Hill* Todd Matuszewicz Metropolitan State University of Denver, De Regan University of Colorado at Denver. Eric Ball and April Hill Department of Chemistry, Metropoli tan State University of Denver, 3058 Science Building P.O. Box 173362. Campus Box 5 2. Denver, CO 80217 3362 (USA). Room 3057. ABSTRACT This research establishes a methodology for extracting fatty acid residue from archaeological pottery sherds in order to verify the use of the vessels and corroborate theories of seasonal migration through dietary habits. Pottery sherds from unidentified sections of the vessels were powdered and washed in solvent to withdraw the embedded fatty residue. Fatty acids are present as free fatty acids or tri, di or monoacylglycerols in all natural animal or plant food sources. ( Table 1 ) Table 1: Fatty acid com position (%) of free fatty acids, mono,di and triacylglycerols of the tropical plant Mabea fistulifera 1 Through various procedures, these fats were mechanically and chemically prepared for analysis U sing computer software the residues were broadly identified through mass spectra fractional analyses and reference to a spectra library Further confirmation was achieved through a comparison to lipid ratio decomposition bi plot data and archaeological speculation. The residues were tentatively identifie d as having originated from a small mammal and a mix of seeds, nuts and berries. To more precisely identify the source substance, further investigation into modern source materials and techniques for isolating proteins, biomarkers and isotopes will need to be studied and developed. INTRODUCTION There are several approaches accessible to investigate the residue in archaeological pottery and determine source material. The choice of lipids as a n analyte wa s based partly on the limit ed financial and analytical instrumentation resources of the project but also on the relative stability of lipid compounds, the ease of analysis ; and the various existing methodologies. 2 While organic residue analysis is somewhat novel in the field of archaeology, it is clear that the interdisciplinary efforts of archaeologists and chemists will yield an increasingly wide array of results 3 One obstacle that is consistent in this field of study is the preponderance of papers concentrating on the examination of results while methodologies are brief and imprecise The re is scant effort to collaborate between investigators to cross che ck methodologies or ve rify the efficacy of procedures. 4 The decomposition of source material and the ability to reverse engineer fatty acids from residues is one of the vexing problems of archaeo chemical analysis. Various initial

PAGE 3

Matuszewicz, Regan Ball and Hill 2 materials can be manipulated by human as well as natural forces; see Figure 1. Figure 1 Schematic of the origin of the organic remains of natural products potentially preserved in archaeological artifacts: they may be from plant or animal origin and may have been used either as raw materials (above) or after various transformations such as heating or fermentation (below). The symbol indicates a fossil origin. 5 The process of simulating oxidation of fatty acids ( Figure 2 ) and experimentally recreating residues in ceramic matrices was also investigated in this paper. The issue of food degradation has relevance in food science as well as in paleontology 6, 7 A lmost all the work has been done exclusively in the former. However, the timescale for archaeological decay is much larger than those for food science and the factors leading to degradation and rancidification more variable. 8 10 This paper will explicitly outline techniques for the destructive pottery analysis of lipid residue in archaeological samples 11 13 Although research has begun on other ancient artifacts such as fire cracked rock or abdominal content from gravesite specific soil sa mples, this paper is limited to ancient ceramics. 14,15 Variou s m ethods of extraction, esterification, derivatization and sily l ation were explored 1 6 18 t his analysis is from these studies and could be an initial attempt to assemble a methodological database for the st andardization of sample preparation practices modeled after spec tral databases such as the S pectral Database for Organic Compound ( SDBS ) Figure 1 Hydrolytic pathway for the transformation of triacylglycerols To free fatty acids. 19 MATERIALS and METHODS Pottery Samples Samples were obtained from four different sources. For the surrogate preparations commercial pottery was initially used 1 0 however, s ubsequent surrogates replicated traditional clay matrixes and manufacturing techniques. Field sample sherds obtained from Dr. D. Hill, and authentic reproductions from the Swink Pottery Studio were employed. Half of the samples were used raw and the other half were ashed at 450 C for 17 hours As shown in Figure 3, t he ashing changed the inner composition of the samples Organic Residues

PAGE 4

Matuszewicz, Regan Ball and Hill 3 Figure 3. A) A raw Swink archaeo pottery sherd. B) An ashed Swink archaeo pottery sherd heated at 450 o C for 17 hours. Of the four archaeological samples only one had any identifying marks (FS#1082.1, Bg#1082, Proj.2033, LA182820, 10 28 2016). Oil Samples Four different oils were used as initial food source material These consisted of avocado, olive, coconut oils and lard. All the oils were obtained commercially and used as received. The samples were stored at room temperature. The coconut and lard were m elted at 6 0 C before applying and sampl ing Sample Preparation Sherds, prepared as either ashed or raw, were coated with the various oils. The first set or group, of sherds contained no additional oil and was used as a blank The other sherds were saturated in oil. Due to the difference in porosity between the samples, the amount of oil applied varied. All of the samples were coated until an even absorption of oil was observed. The second set remained un heated The third set of sherds were heated in an oven at 250 C for 1 h ou r. The fourth set of sherds heated 1 hour at 250 C, cooled 1 hour to room temperature, and heated 1 hour at 250 C The fifth set of sherds were heated 1hour at 250 C cooled 1hour to room temperature, and cooked overnight at 250 C The sixth set of sherds cooked overnight at 75 C ( Table 2 ) Table 2 : The preparation of various sample sets for degradation simulations Chemicals Solvents used in the extraction and preparation of the oils from the prepared sherds included dichloromethane chloroform, methanol, hydrochloric acid petroleum ether and iso octane ( 2,2,4 trimethylpentane) All solvents were purchased from Fischer Scie ntific (Waltham, MA, USA) except t he petroleum ether which was purchased from Mallinckrodt ( Sta ines Upon Thames, Surrey UK ) The 0 .5 N anhydrous hydrochloric acid and N,O b is(trimethylsilyl) trifluoro acetamide ( BSTFA) + Trimethyl chlorosilane (TMCS), 99:1 a column evaluation FAME mix and a C8 C24 FAME standard mix were purchased from Supelco (Bellefonte, PA, USA) Extraction Lipids were extracted using a combination of methodologies 20 22 Samples 1 cm 2 were crushed with a mortar and pestle. A 400 mg portion of each powdered sample was placed in a test tube and mixed with 5 mL of a 2:1, ( v:v. ) mixture of chloroform and methanol or dichloromethane and methanol Samples were sonicated for 10 min utes and centrifuged for 10 minutes. The super natant

PAGE 5

Matuszewicz, Regan Ball and Hill 4 was syringe filtered through a Thermo Scientific 0.20 m PDVF membrane filter and transferred to a vacuum centrifuge tube. The extraction process was repeated with a 15 sec ond vortexing of the sample/ solvent solution before sonication and centrifuging. The 10 m L aliquot of lipid/solvent was vacuum centrifuged 120 minutes at 37 o C until less than 1.0 mL of analyte remained. Samples were transferred to a screw top glass vial with a Teflon s eptum cap and flushed with nitrogen (N 2 ) before storage in 20 C freezer. Free Fatty Acid Derivatization Several 4 00 L aliquots of the total lipid extract /solvent were vacuum centrifuged until dry. The dried lipid s were transferred into Teflon septum screwtop glass vial s. Fatty acid methyl esters (FAMEs ) were prepared and derivatized through two methods ; see Figure 4 The first by adding 3.0 mL of 0.5 N anhydrous hydrochloric acid in methanol and heating for 60 min at 70 C. The second used 4.1 ml of 0.5 M methanolic HC l and was heated for 60 minutes at 68 o C in a heating block. These methods ensured that any f atty acids in the sample still attached to the glycerol backbone as tri di or monacylglycerols (TAGs, DAGs, or MAGs ) were detached and converted to FAMEs (esterification) After cooling to room temperature, the FAME s were transferred to a test tube with 3 x 1 m L of water. The FAMEs were retrieve d with 2.5 mL of petroleum ether. A 1000 L Finn Pipette II was used to remov e the upp er solvent layer and transfer it to a crimp top GC vial. The solution was dried in a vacuum centrifuge ( 15 min.) at 37 C to remove the solvent The dried FAMEs were dissolved with 75 L of iso octane and transferred to a GC vial with a 100 L conical insert. Figure 4. The mole cular structure of a fatty acid methyl ester FAME and the mechanism for esterification. 23 Silylation Preparation of BSTFA and TMCS Derivatives A 100 L aliquot of the total lipid extract s olution was transferred to a Teflon septum screw top 4 mLvial Trimethylsilyl (TMS) derivatives were prepared by adding 70 L N,O BSFTA and TMCS 99:1, ( v:v ) and heating in a heating block for 30 min utes at 70 C. After cooling to room temperature, 75 L of the derivative was transferred to a GC vial with a 100 L conical insert. GC /MS Analysis The GC analysis was performed on a Hewlett Packard (HP) 6890 gas chromatograph fitted with a Hewlett Packa rd 5973 mass selective detector, a 7683 series auto injector and connected to a personal computer equipped with ChemStation software Samples were run through a Zebron ZB5 HT column with a (5% Phenyl) methylpolysiloxane phase Linoleic Methyl Ester

PAGE 6

Matuszewicz, Regan Ball and Hill 5 specifically designed for the high quality separation of fatty acid methyl esters (FAMEs) The ZB5 H is a sil ica fused capillary ( 30 m x 0.32 mm I.D. 25 ; Phenomenex, Torrance, CA,USA). An autosampler injected a 3 volume using a splitless injecti on system. Helium was the carrier gas with a column flow rate of 1.0 mL /min. Two different GC programs were run. The first method involved an inlet temperature of 250 o C and t he ini tial column temperature was 100 C; thi s was increased at a rate of 12 C per minute to a final temperature of 325 C, which was held for 12 minutes. The second method involved the same column and instrument but had a slower ramp The initial temperature was 140 o C rising to 25 0 o C at 6 o C/minute with a 5 minute hold at the end. The c hromatogram peaks were integrated using HP Chem Station software .The mass spectr a were compared against the National Institute of Standards and Technology ( NIST ) Mass Spectral library (1998) Solvents, chemicals, pottery stock and the column were checked for impurities by running blank samples before during and after each group of samples Results D erivatizations of Neat O ils The first attempt at derivatiz a tion was the use of the neat oils avocado, olive, coconut and lard. Random quantities of the fats, between 100 and 800 mg, were heated at 60 o C for 4 hours 45 minutes with 3 mL of methanolic HCl in a Teflon septum screw top vial The results of the FAME abundances correlates to the lipid ratios of the raw oils see Appendix A The lipids follow the nomenclature system of C16:1 where C is carbon, the first number is numbers of carbons in the molecule, colon, and the second number i s number of double bonds or degrees of unsaturation, see Table 3 for ratios. The extraction and derivatization techniques were not subject to analysis as the neat samples provided abundances too large to assess extraction and derivatization. Table 3 Th e GC/MS analysis of neat oils using fast temperature ramping in the column parameters. The four oils: avocado, coconut, olive and lard, were derivatized and analyzed and the results were compared against literature data. Surrogates and Degradations Surrogates were prepared using pottery sherds obtained from three sou r ces: Home Depot ,Inc. Swink Pottery Studio, and field samples from Costa Rica obtained from Dr. D. Hill which registered no lipid residue on previous analysis. Half of the samples were o C for 24 hours to destroy any organic residues. Lipids begin to rapidly denigrate at 80 C and with higher temperatures any incidental contamination is removed. This process creates a clean sherd that is the platform for oils to be artificially degradated in an attempt to model

PAGE 7

Matuszewicz, Regan Ball and Hill 6 environmental degradation. The other half were left u nheated so as not to modify the traditional unfired carbon core of archaeological pottery. The four neat oils were spread on three of the ashed sherds o ne from each of the suppliers. The first group of three sherds were left as is with raw oil. The second group of three sherds were heated at 25 0 o C for 1 hour. The third group of three sherds were heated 1 hour at 250 o C, allowed to cool to room temperature for 4 ho urs and then heated another hour at 250 o C. Although each of the oils provided results, olive oil became the preferred oil of investigation for practical re a sons of simplification. The results of this inquiry are displayed in T able 4 Table 4 Olive oil on two ashed sherds from a previous investigation in Costa Rica. The uncooked (raw) samples exhibit traditional lipid ratios for olive oil. The cooked samples have no lipid residue remaining in them. Although four oils were investigated, only the olive was utilized as a standard. The chromatogram s are of olive oil uncooked and heated for 2 hours at 250 o C are exhibited i n Figure 5 and Appendix A A second degradation method was employed on the a rchaeo pottery sherds from Swink Pottery As expected, t he ashing changed the inner matrix of the sherds ; see Figure 3. Sherds were analyzed without lard, w ith lar d but uncooked and with lard and cook ed at 75 o C for 48 hours ; see Table 5. Figure 5 The ch romatogram for the fast ramp GC program of methyl ester derivatized olive oil uncooked on an ashed surrogate pottery sample. Table 5 Total Lipid Extraction (TLE) of various fatty acids in surrogate pottery samples. The top group is the raw sherds with no lard, with lard but uncooked, and with lard and cooked at 75 o C for 48 hours. The loss of abundance is shown as a percentage loss. The lower group is the ashed sherds with no lard, with lard but uncooked and with lard and cooked at 75 o C for 48 hours. The loss of abundance is shown as a percentage loss. The difference in degradation is shown as a ratio. The raw degrades faster and mono unsaturated oils (MUFAs) degrade faster than saturated oils (SFAs).

PAGE 8

Matuszewicz, Regan Ball and Hill 7 Extraction and Derivatization Method Several different extraction and derivatization techniques were investigated over the course of the project. The multiplicity of variations is beyond the scope of this paper. In brief, it was concluded that a chloroform methanol solvent 2:1 (by vol ume) was the most practical and effective at extracting lipids The use of methanolic HCl to derivatize the fatty acids into FAMEs provided the clearest and most useful dat a. None of the other methodologies were effective and the unusable data wa s left out of this an alysis. The final extraction and derivation procedure is : The Extraction of Lipid Residue 1. Crush sample with mortar and pestle and transfer 400 mg to a test tube 2. Powdered sampl e is mixed with a 2:1 (by volume)mixture o f chloroform and methanol ( 5 ml). 3. Ultrsonicated ( 10 min). 4. Centrifuged (10 min). 5. Filter solids from solvents through syringe filter into a vacuum centrifuge test tube. 6. Repeat steps 2 5 7. Vacuum centrifuge at 37 C until less than 1.5mL of analyte remains ( Sample can be transferred to a Teflon screw top vial, flushed with nitrogen and stored in 20 C freezer ) Preparation of Fatty Acid Methyl Esters FAMES 8. A 400 in a GC vial in a vacuum centrifuge tube. Vacuum centr ifuge at 37 C until dry. Approximately 15 min. 9. Fames are prepared by treating the dry lipid with 4.0 mL of 0.5 M methanolic HCl. 10. Heat in h eating block in capped Teflon screw top vials for 60 min at 68 C 11. Vortex for 15 sec. 12. After cooling to room temp, transfer to test tube using 3.4 mL of ultrapure water. 13. FAMES are recovered using petroleum ether (2.5 mL). 14. Transfer the top layer to a GC vial using a 1000 autopipette 15. Solvent is removed using heat V acuum centrifuge at 37 C until dry) 16. FAMES are dis octane. 17. Transfer to GC vial with conical insert. An illustrated detailed procedure is outlined in A ppendix B GC /MS P rograms Two separate GC/MS programs were used. The fast temperature ramp was the first method and the slow ramp was the second. The re tention times for various FAMEs are listed in Tables 6 and 7 respectively T he slow ramp chromatogram of fatty acid met hyl ester standard s is shown in Figure 6. Table 6 Retention times of fatty acid methyl esters (FAMEs) using a fast temperature ramp GC/MS program. Table 7 Retention times of fatty acid methyl esters (FAMEs) using a slow temperature ramp GC/MS program.

PAGE 9

Matuszewicz, Regan Ball and Hill 8 Figure 6 The chromatogram for the slow ramp GC program of fatty acid methyl esters. Archaeological Samples Four archaeological samples and the dirt matrix from which they were excavated were analyzed. Two of the archaeological samples showed no significant difference between the total lipids o f the soil and the sherd s Tho se two samples are not included in this analysis. The total lipid extracts of two other specimens, pot sherd s # 1 and # 4 are in Table 8 P ot sherd #1 did not have sufficient residue to warrant analysis. Table 8 Total Lipid Extraction (TLE) from two archaeological samples, pot sherd #1 and pot sherd #4. The analysis used a slow ramp GC/Ms program. Abundances were measured in peak area. Sherds #2 and #3 had negligible TLE. Discussion Extraction and Derivatizations Method s other than the one from the methods section, which is further detailed in Appendix B did not produce any useful results. The chromatogram in Appendix A shows the results of one of the silylation derivatizations. Given the considera ble time involved in adapting the esterification derivatization, a silylation methodology is beyond the prevue of future research in this lab. The extraction method simplified previous methods, eliminating the need for expensive chemicals (0.5 N anhydrous hydrochloric acid) in step 9 and the use of a nitrogen glove box in steps 7, 8 and 15 The method did leave question s about the efficacy in regards to the identification of medium chain FAMEs and very long chain FAMEs. It is not clear whether this methodol ogy can effectively ext ract and derivatize lipids medium and very long chain compounds. Surrogates and Degradations The Swink a rchaeo pottery sherds were not compared against the Home Depot sherds or previously negated field samples. The Swink sherds were analyzed using a different degradation method than the other two The second group of samples returned no viable data due to overly high temperatures in the degradation process. Without a formal analysis, it is clear from Table 4 and Appendix A that t he degradation procedure of 250 o C destroyed any lipids. In the archaeo sherds, t here w as considerable difference between the degradation of fatty acids in the raw sherds versus the 'ashed' sherds ( Table 5 ). Lipids C14:0, C16:0 and C18:0 degraded 1.81, 1.56 and 1.75 times faster, respectively, in the raw sherd than in the ashed sherd. Unquestionably, the composition of the surrogate sherds affects the decomposition rate of lipids. The mono unsaturated fatty acids were more unst able than the saturated lipids ( Table 5 ) The greater the number of double bonds in a fatty acid, the greater the oxidation

PAGE 10

Matuszewicz, Regan Ball and Hill 9 susceptibility due to the addition of more methylene interrupted carbon reaction sites This trend is in agreement with previous studies ( Figure 7 ). C16:1 and poly unsaturated fatty acids were absent in the degraded samples. This suggests a necessary adjustment to future method s Figure 7 Simulated degradation for four fatty acid ratio s over 100 decomposition steps. 24 GC /MS Programs By comparing the tw o separate GC/MS temperature programs the slow ramp allows for the identification of the medium chain fatty acids, C8:0 C12:0, and the very long chain C22:0 C24:0. Although the fast ramp method seems to indica te an advantage in the i dentification of odd long chains, it is difficult to compare due to the fact that those lipids were not present in the standards used for the slow ramp analysis ( Tables 6 and 7 ). The slow ramp is preferable as it allows for clearer results through better resolution of the peaks. Archaeological Samples Only one of the archaeological samples provided sufficient lipid residue for analysis, number 4 B y using a lipid ratio analysis (Table 9 ) and comparing it against the lipid biplots from literature in Figures 8 and 9 the composition appears to have originated from terrestrial mammal or seeds. These biplots were generated from previous research in the western regions of Canada. 21 By drawing a vertical line for the numerical ratio of t he lipids C16:1/C18:1 and a horizontal line for the ratio of (C15:0 + C17:0)/C18:0, the intersection designates the origin source. In Figure 8 that would be roots or seeds. Figure 9 is another example of a biplot ratio graph using different lipids. Table 9. Analysis of long chain fatty acids (LCFAs) in archaeological pottery sherd #4. Figure 8 Biplot of two conservative fatty acid ratios for modern food products (ellipses represent subjective estimations of the range of values for each food group). The vertical line represents the known lipid ratio of C16:1:C18:1. 17

PAGE 11

Matuszewicz, Regan Ball and Hill 10 Figure 9 Biplot of two additional conservative fatty acid ratios for modern food products (ellipses represent subjective estimations of the range of values for each food group). The vertical line represents the known lipid ratio of C16:1:C18:1. 17 Further lipid ratio analysis confirms the same two food groups sources ( Table 10 ). Table 10 Distinguishing food types by fatty acid ratios (after Eerkens 2001,106). Some seeds and nuts have values well over 4, making this a worthwhile ratio. The evide nce from sherd #4 indicates that the source material was likely mammal and seeds and nuts. The last analysis of the lipids in sherd number 4 is a straight percentage of various lipids detected in the analysis The results indicate the lipid sources to be from terrestrial mammal and a mix of seeds, nuts and berries ( Table 11 ) Table 11 Analysis of individual lipid percentages of the Total Lipid Extract (TLE) from pottery sherd number 4. These percentages confirm the presence of mammal and a mix of seeds, nuts and berries. 10 Further Research This project marks the initiation of exploration in to the field of archaeo chemistry. Future experiments can provide more information on the extra ction and derivitization of medium and very long chain fatty acids. More methods for the degradation of lipids, including light and moisture and new techniques of heating samples should be investigated A database of regional specific food sources and thei r lipid composition after preparation is needed to more precisely identify origin materials. The investigation of biomarkers, isotopic relationships, yeast residues, proteins and electon microscopy are all potential addition s to the lipid research. 25 2 8

PAGE 12

Matuszewicz, Regan Ball and Hill 11 CONCLUSION With a simplified extraction/derivatization technique, the use of archaeo pottery samples as surrogates, a database of regional source materials, the lipid composition of those source materials and an artificial degradation model, the possibility of historical insights into dietary, medicinal and cultural practices through chemical analysis is unlimited. 15,18,27 This paper has outlined a simple and specific methodology that uses inexpensive chemicals and equipment and can be easily learned by a novice chemist. The GC temperature program outlined in this work produces clear peak resolution. With these standardized techniques the collection of greater source origin material lipid compositions is possible. Combined with conti nued investigation into lipid degradation, the analyses of archaeochemistry will become more accurate. Acknowledgements We would like to thank David Hill for the opportunity to work on the archaeological food residue project, t he Undergraduate Research Department of Metropolitan State University of Denver provided funding for the GC/MS analysis and Clint Swink for providing traditional Anasazi pottery for surrogate preparation. Ref erences 1. Scientific Electronic Online Library http://www.scielo.br/scielo.php?scrip t=sci_arttext&pid=S0100 40422008000300002 (accessed May 11, 2017). 2. Murphy, R.C. Mass Spectrometry of Lipids. Handbook of Lipid Research Volume 7. 1993 New York: Plenum Press. 3. Eerkens, J.W. and Barnard H. Introduction. The Theory and Practice of Archaeological Residue Analysis. Introduction 2007 Oxford, England: Archaeopress. 4. Barnard, H.; Dooley A.N. and Fauli, K.F. An Introduction to Archaeological Lipid Analysis by Combined Gas Chromatography Mass Spectrometry (GC/MS). The Theory and Practice of Archaeological Residue Analysis. Chapter Five. 2007 Oxford, England: Archaeopress. 5. Regert, M.; Garnier, N.; Decavallas, O.; Cren Oliv, C. C. and Rolando, C. Structural characterization of lipid constituents from natural substances preserved in archaeological environments. Meas. Sci. Technol. 2003, 14, 1620 1630. 6. Collins, M.J.; Nielsen Marsh C.M. ; Hiller, V; S mith, C.I.; Roberts, J.P. ;

PAGE 13

Matuszewicz, Regan Ball and Hill 12 Prigodich R.V. ; Wess, T.J. ; Csapo, J. ; Millard A.R. and Turner Walker G. The Survival of Organic Matter in Bone. A Review. Archaeometry 2002 44,3: 383 394. 7. Frankel, E.N. Lipid Oxidation. 1998 Ayr: Oily Press. 8. Evershed R.P. Biomolecular Archaeology and Lipids. World Archaeology 1993 25(1):74 93. 9. Malainey, M. E. Lipid Residue Analysis. Manuals in Archaeological Method, Theory and Technique. Chapter 14. 2011, pp.201 217. New York, Dordrecht, Heidelberg, London: Springer Press. 10. Malainey, M. E.; Przybylski, R. and Sherriff, B. L. The Fatty Acid Composition of Native Food Plants and Animals of Western Canada. J. Archaeol. Sci. 1999, 26, 83 94. 11. Evershed R. P. Biomolecular Analysis by Organic Mass Spectrometry. In Modern Analytical Methods in Art and Archaeology 20 00 edited by E. Ciliberto and G. Spoto, Volume 155: 177 239. Chemical Analysis. John Wiley & Sons, New York. 12. Craig, O. E.; Steele, V. J.; Fischer, A.; Hartz, S.; Andersen, S. H.; Donohoe, P.; Glykou, A.; Saul, H.; Jones, D. M.; Koch, E. and Heron; C. P. A ncient lipids reveal continuity in culinary practices across the transition to agriculture in Northern Europe. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 17910 17915. 13. Shishlina, N.I. The Seasonal Cycle of Grassland use in the Caspian Sea Steppe during the Bronze Age. A New Approach to an Old Problem. European Journal of Archaeology 2001 4,3: 346 366. 14. Berg, G.E. Last meals. Recovering Abdominal Content from Skeletonized Remains. Journal of Archaeological Science 2002 29: 1349 1365. 15. Quigg, M.J.; Malainey, M.E. ; Przybylski R. and Monks G. No Bones about it. Using Lipid Analysis of Burned Rock and Groundstone Residue to examine Late Archaic Subsistence Practices in South Texas Plains Anthropologist 2001 46. 283 303. 16. Aveldano, M.I., VanRollins M. and Horrocks L.A. Seperation and Quantation of Free Fatty Acids and Fatty Acid Methyl Esters by Reverse Phase High Pressure Liquid Chromatography. Journal of Lipid Research 1983 24:83 93 17. Eerkens, J.W. Organic Residue Analysis and the Decomposition of Fatty Ac ids in Ancient Potsherds. The Theory and Practice of Archaeological Residue Analysis. Chapter Eight 2007 Introduction. Oxford, England: Archaeopress. 18. Patrick, M.; de Konig, A. J. and Smith, A. B. Gas Liquid Chromatographic Analysis of Fatty Acids in Food Residues from Ceramics Found in the Southwestern Cape, South Africa. Archaeometry 1985, 27(2): 231 236. 19. Frankel, E. N. Recent Advances in Lipid Oxidation. J. Sci. Food Agric. 1991, 54, 495 511. 20. Eerkens,J.W., British Archaeological Reports International S eries 1650 pp.42 60. Oxford, UK. 21. Folch, J.; Lees, M.; and Sloane Stanley, G. H. A simple method for the isolation and purification of lipid extracts from

PAGE 14

Matuszewicz, Regan Ball and Hill 13 brain tissue. J. of Bio.l Chem. 1957, 191:833. 22. Malainey, M. E.; Malisza, K.L.; Przybylski R. and Monks G. The Key to Identifying Archaeological Fatty Acid Residues. Paper presented at the 34 th Annual Meeting of the Canadian Archaeological Association Banff, Alberta, May 2001 23. Evershed R. P.; Dudd, M.; Copley, M.S.; Berstan, R.; Stott, A.W.; Mot tram, H.; Buckley S.A. and Crossman Z. Chemistry of Archaeological Animal Fats. Accounts of Chemical Research 2002 35:660 668. 24. DeMan, J .M. Chemical and Physical Properties of Fatty Acids. In C.K. Chow (ed.) Fatty acids in foods and their health implications. 2008 CRC PressBoca Raton., pp.17 45. 25. Barnard H. and Eerkens J.W. Abstracts for the Symposium on Theory and Practice of Archaeological Residue Analysis at the 70th Annual Meeting of the Society for American Archaeology Salt Lake City (Utah), March 2005 26. Budja, M. Neolithic pottery and the biomolecular archaeology of lipids. Documenta Praehistorica 2014 41 : 196. 27. Evershed R. P.; Heron C. and Goad L.J. Analysis of Organic Residues of Archaeological Origin by High Temperature Gas Chromatography and Gas Chromatography Mass Spectroscopy. Analyst 1990 115:1339 1342. 28. Wang, J.; Liu, L.; Ball, T.; Yu, L.; Li, Y. and Xing, F. Revealing a 5,000 y old beer recipe in China. Proceedings of the National Academy of Sciences 2016 113 6444 6448.

PAGE 15

Matuszewicz, Regan Ball and Hill 14 Appendix A: Figure 10 The chromatogra m for the fast ramp GC program of methyl ester derivatized neat avocado oil. Figure 11 The chromatogra m for the fast ramp GC program of methyl ester derivatized neat coconut oil.

PAGE 16

Matuszewicz, Regan Ball and Hill 15 Figure 12 The chromatogra m for the fast ramp GC program of methyl ester derivatized neat olive oil. Figure 13 The chromatogra m for the fast ramp GC program of methyl ester derivatized neat lard.

PAGE 17

Matuszewicz, Regan Ball and Hill 16 Figure 14 The chromatogram for the fast ramp GC program of methyl ester derivatized olive oil cooked on an ashed surrogate pottery sample at 250 o C for 2 hours. Figure 15 The chromatogram for the slow ramp GC program of fatty acid methyl esters.

PAGE 18

Matuszewicz, Regan Ball and Hill 17 Figure 16 The chromatogram for the fast ramp GC program of silylation derivatized olive oil uncooked on an ashed surrogate pottery sample.

PAGE 19

Matuszewicz, Regan Ball and Hill 18 Appendix B: Extraction and Derivatization Method 1. Crush sample with mortar and pestle and transfer to a test tube (13 x 100mm). 2. Powdered sampl e is mixed with a 2:1 mixture v:v o f chloroform and methanol (5 ml). 3. Ultrsonicated ( 10 min). 4. Centrifuged (10 min).

PAGE 20

Matuszewicz, Regan Ball and Hill 19 5. Filter solids from solvents through syringe filter (0.2 m) into a vacuum centrifuge test tube (15 x 85mm lime glass). 6. Repeat steps 2 5 7. Vacuum centrifuge at 37C until less than 1.5mL of analyte remains ( 2 hours).

PAGE 21

Matuszewicz, Regan Ball and Hill 20 ( Sample can be transferred to a Teflon screw top vial, flushed with nitrogen and stored in )

PAGE 22

Matuszewicz, Regan Ball and Hill 21 Preparation of Fatty Acid Methyl Esters FAMES 8. in a GC vial which is placed in a vacuum centrifuge tube. V acuum centrifuge at 37C until dry 15minutes.

PAGE 23

Matuszewicz, Regan Ball and Hill 22

PAGE 24

Matuszewicz, Regan Ball and Hill 23 9. Fames are prepared by treating the dry lipid with 4.0 mL of 0.5 N anhydrous hydrochloric acid in methanol or 0.5 M HCl/MeOH in a heating block capped Teflon screw top vials (68C; 60min). 10. Vortex 15 seconds.

PAGE 25

Matuszewicz, Regan Ball and Hill 24 11. After cooling to room temp, transfer to test tube. 3.4mL of ultrapure water is added.

PAGE 26

Matuszewicz, Regan Ball and Hill 25

PAGE 27

Matuszewicz, Regan Ball and Hill 26 12. FAMES are recover ed using 2.5mL petroleum ether Using a 1000 L autopipette pull out 2000 L, the top layer,t ransfer to a GC vial.

PAGE 28

Matuszewicz, Regan Ball and Hill 27 13. Solvent is removed using heat Place GC vial in vacuum centrifuge tube. Heat at 37C until dry 15 minutes.

PAGE 29

Matuszewicz, Regan Ball and Hill 28

PAGE 30

Matuszewicz, Regan Ball and Hill 29

PAGE 31

Matuszewicz, Regan Ball and Hill 30 14. FAMES are dis octane added to dry GC vials using a 100 L autopipette.

PAGE 32

Matuszewicz, Regan Ball and Hill 31 15. Transfer to GC vial with conical insert using the same autopipette.

PAGE 33

Matuszewicz, Regan Ball and Hill 32

PAGE 34

Matuszewicz, Regan Ball and Hill 33