Roman Pozzolana and mortar datingC: Applications

and experiments29 samples

CI

2 inconclusivesamples

CI-CII

19 conclusivesamples

CIV

7 samplesfulfill less secureCriteria III and IV

CIII

1 inconclusivesample

Secondfractionsagree in 12 samples from Trajan’sMarket and Ostia

B: No age controlA: Age control

0

100

200

300

400

500

600

1 2 3 4 5 6 7 8

1

2

3

4

5

6

7

8

All mor-tars

Limemor-tars

Pozzo-lana

Lime lumps

H3PO4 HCl Clas-sical

Diagram of all mortar samples analyzed

Type Hydrolysis Chronology

Medi-eval

This poster focuses on recent methodological development in dating mortar, exspecially on classical archaeology and the dating

of lime lumps embedded in the mortar

Mortar is not an organic material. Yet the chemical process in the hardening in principle makes it into an ideal matrix for 14C dating. To make mortar, limestone has to be heated up to at least 900 ºC. After the carbon dioxide has been released in the process, calcium oxide (un-slaked lime) is mixed with water and aggregate, usually sand. In the hardening process the slaked lime reacts with atmospheric carbon dioxide, and calcium carbonate is produced. Thus the mortar absorbs the carbon dioxide from the atmosphere and thereafter behaves as if it were organic. Unburned limestone in the mixture can contaminate the result, yielding dates too old.

Diagram displaying mortar samples analyzed. By now the databank covers results from more than 550 analyses of some 470 individual samples. Some of the samples were analyzed repeatedly, in different laboratories. The samples include different sorts of mortar, and they cover both Classical and Medieval archaeology.

Recent Advances in Radiocarbon Dating of Ancient Roman Mortar and Concrete

Lime lumps embedded in mortar

In the process mortar is more easily soluble than limestone, thussthe first fractions should be free from contamination. Therefore, to illustrate the entire dissolution process, the gas-flows interrupted at several stages create a sequence of CO2 age profiles. The contamination is shown towards the end of the profile.

21 3 4 5

5 CO2-fractions

0 100CO2 in the sample (%)

14C-

age

(BP)

To avoid the effects of contamination different types of separation are needed. At first a mechanical separation with gentle crushing of the sample. The grains are then sieved into 39/45-75 microns for further analysis. In the following chemical separation 85% of phosphoric acid is poured over the sample, creating a gas of carbon dioxide.

LIQUIDNITROGEN

CO2 FRACTIONS

COLD TRAP

SAMPLE

PRESSUREGAUGE

VACUUM PUMP

Åsa Ringbom, Åbo Akademi University, FinlandJohn R. Hale, University of Louisville, Kentucky

Torre delle Milizie, Rome, The result of Rome 007Li, (Li=lime lump) as compared to analysis of Rome 007 from different AMS laboratories (Aarhus, Denmark, Oxford UK, Tucson Arizona). The lime lump analysis supports our earlier interpretations - that the top of Torre delle Milizie was erected towards the end of the 13th century.

Santa Costanza, at Via Nomentana, Rome. The age of construction is debated. It was built as a mausoleum either by Constantina, daughter of Constantine the Great, before 350, or by her brother–in–law Julian the Apostate some 10-12 years later. Mortar samples were analyzed in Oxford, Tucson and Aarhus all forming confusing age profiles. The lime lump Rome 042 was analyzed twice, resulting in an almost horizontal age-profile, which is easy to interpret. But because the calibration curve wiggles at this point the result is not helpful enough in pinpointing the chronology. We can never defeat the irregularities of the calibration curve.

Conclusions: a) Medieval lime mortars in Scaandinavia and Classical mortars based on marble (Torre de Palma) have yielded promising results. b) Hydraulic Pozzolana mortars have been less successful.c) Pozzolana mortars from Pompeii and Herculaneum cannot be dated, regardless of whether bulk mortars or lime lumps embedded in the mortars are used. d) Lime lumps embedded in the mortar have shown a great deal of success, both in lime - and pozzolana mortars. Lime lumps generally contain very little contamination from unburnt limestone, thus fewer CO2 fractions are needed, thus cheaper. e) Above all, for pozzolana mortars analyzing lime lumps may be the obvious solution.

Why mortar? The aim of archaeological excavations is often to establish the chronology of a site. Mortar, lime mortar or pozzolana mortar, is often the only datable material available, in abundance, from all the different building stages. Differently from other datable materials like dendrochronology and 14C of organic materials, and differently from archaeological artifacts such as coins and ceramics, mortar analysis dates the actual time of the construction, or when the mortar hardened. It would therefore be helpful to use mortar dating on a routine basis in archaeological excavations or in a process of conservation.

The Roman Villa of Torre de Palma, Portugal, mortar analyzed in 1997-2000, was our first encounter with Classical archaeology. Out of a total of 64 samples analyzed 18 yielded Criterion I results. Independently of other materials and methods these 18 samples could indicate the date of 12 out of 21 structures, thus marking the outlines of the chronology of the site. The earliest structure was the temple in the East Court (70-170AD), and the latest are the apse 3 of the Basilica (580-630AD) and the large font of the Baptistery (570-690AD).

Reliability criteria for the interpretation of the age profiles

Criterion I: when analyzed in two to five CO2 fractions and the results from the two first CO2 fractions agree Criterion II: when there is mutual agreement between the dates of the first CO2 fractions in a seriesof three or more samples from one single building.Criterion I and II yield a conclusive date independently of age control.Dating medieval lime mortars in Scandinavia has

yielded promising results.More than 96% if all the results agree with age control from other dating methods, as shown from the Åland Islands between Finland and Sweden

Dating pozzolana mortars, however, was problematic. Only some 50% of all samples could be considered successful, and the interpretation of the CO2 age profiles was seen as complicated.

A Mortar Dating Workshop in Åland, 2006: Fiona Brock, John R. Hale, Jan Heinemeier, Åsa Ringbom, Alf Lindroos and Pia Sonck-Koota.

The International Mortar Dating Project 1994 - Map of mortar dating sites, those in red mark lime mortars, those in green pozzolana mortars.

200 4000 km

N

CII

Conclusive results 45

Inconclusive results 26

CI

Not agreeing 4

Mutually agreeing results 75

CII

CIII CIV

Age control, 79 samples No Age control, 71 samples

Dendro Och 14CW

14C W

Dendro and 14C W

Dendro CI

CI-CII

CII

CI-CII

CI

CIII, Too few samples per building- unit. Age- Profiles cannot be interpreted

The mortar dating group, Lynne Lancaster, Jan Heinemeier, Åsa Ringbom and Alf Lindroos outside the Colosseum, where successful results meet the demands of Criterion I and Criterion and II, and the calibrated result agrees with written sources.

1800

1900

2000

2100

2200

2300

2400

2500

2600

2700

2800

2900

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1F

C-14 age BP Colosseum

003, 76-150 m

001, 46-75 m

002, 46-75 m

Heavilycontaminated

No re-crystallizationFirst fractions convergeand are correct

Atmospheric data from Reimer et al (2004);OxCal v3.10 Bronk Ramsey (2005); cub r:5 sd:12 prob usp[chron]

200CalBC 100CalBC CalBC/CalAD 100CalAD 200CalAD 300CalAD

Calibrated date

1800BP

1900BP

2000BP

2100BP

noitanimreted nobracoida

R

The Colosseum : 1940±17BP 68.2% probability 25AD (13.4%) 40AD 45AD (54.8%) 80AD 95.4% probability 10AD (86.5%) 90AD 100AD ( 8.9%) 130AD

Why lime lumps?

Lime lumps embedded in the mortar are highly interesting in this context. There may still be clods of quicklime in the putty, which become slaked and carbonated possibly already before the aggregate is added. These form aggregate free white lumps in the mortar. In principle therefore well-burnt lime lumps should be free from contamination.

CO2 CaCO3limestone

CaOunslaked lime

Ca(OH)2slaked lime

+ H2Owatercarbon-

dioxide

Statistics of results from 32 liumps embedded in the mortar. 25 lime lumps were identified in lime mortars, whereas 7 were embedded in pozzolana mortars. Of these 32 samples, 7 had been exposed to fire. The majority comes from medieval Scandinavia.

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7

1

2

3

4

5

6

7

All limelumps

successrate

Nordicmedieval

Classical Pozzo-lanaRome

Pozzolana Pompeii

Torre dePalma

Current statistics of lime lumps

Terracina, Temple of Anxur. Repeated analysis of a bulk sample of pozzolana mortar (Terracina 001)resulted in two different looking age profiles, which seemed to converge at the first CO2 fractions. A lime lump embedded in the pozzolana confirmed the earlier interpretation.

500

700

900

1100

1300

1500

1700

1900

2100

2300

2500

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

C-1

4 ag

e B

P

F

Rome, Torre delle Milizie

Oxford 76-150 m H3PO4

Sample Rome 007

Tucson 46-75 m H3PO4

Tucson 46-75 m HCl

ÅrhusLime lump 21-150μmH3PO4

Århus 46-75μm H3PO4

CO2CaCO3

CaO Ca(OH)2 +sand

+ H2O+ H2O

14CO2from atmosphere

Samples from Pompeii or

Herculaneum

Delayed hardening

Experiments

Crushed brickssec. opus floors

Too few fractions analyzed

Re-crystallization

with old layersof stuccho

200

45-80

10-22

6-3

Atmospheric data from Reimer et al (2004);OxCal v3.10 Bronk Ramsey (2005); cub r:5 sd:12 prob usp[chron]

Torre de Palma - Dates based on agreement between 1st and 2nd fractions in CO2

600BC 400BC 200BC BC/AD 200AD 400AD 600AD 800AD

Calibrated date

The Temple in East CourtAAR-3933 TP 108 Cat1 1885±35BP South Field WestAAR-5441 TP 186 1683±45BPAAR-5442 TP 188 1750±45BP south west houseAAR-3929 TP 069 1711±35BP The Baptistery, earlier structureAAR-10361 TP 013 1684±35BPAAR-3411 TP 201 1738±55BP Precinct wall AAR-5440 TP 182 1601±35BP The Olive PressAAR-4231-2 TP 083-2 1628±55BP The Basilica, phase 2AAR-3926 TP 023 1535±30BPAAR-4825 TP 143 1554±20BP Basilica, Apse 4AAR-5653 TP 209 1483±30BP North East House AAR-5428 TP 045 1543±35BPAAR-5643-1 TP 178-2 1515±45BP MaosoleumAAR-4228 TP 036-2 1576±45BP Basilica, Apse 3AAR-3927 TP 025 1410±35BPAAR-3928 TP 026 1448±30BPAAR-3414 TP 206 1417±30BP Babtistery, large font, second stageAAR-4822-3 TP 138-3 1393±40BP

North Barn

SOUTH F IELD WEST

Southwest Cemetery

(CW R)

Northwest CemeteryBASILICA

NORTHEAST BUILDING

Garden Quarter

Peristyle House

East Bath

Forge

Atrium House

East Court

South Hall

West Court

GateGate

Gate

Olive Press

Portico House

West Bath

Southwest House A

Southwest House B

(COMPLEX B)

(COMPLEX C)

(COMPLEX R)

South Field

TORRE DE PALMA

Location of Studies Samples

2325

26

1013

138

1434

182

10 m

209

144,219

145

211

147

175

32199139

TORRE DE PALMABasilica (Complex B)Criterion I (black)Criterion II (red)

17

144,1

1800

2000

2200

2400

2600

2800

3000

0 0,2 0,4 0,6 0,8 1

Terracina, temple of Jupiter Anxur

001; 46-75µm001Li; 21-150μm

F

Sample 001

Lime lumpH3PO4, Århus

Bulk mortars

14C

age

BP H3PO4 (Århus)

HCl (Tucson, Arizona)

1300

1400

1500

1600

1700

1800

1900

2000

2100

2200

2300

2400

0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

St Costanza, RomeC-14 age BP

F

Samples

Rome 042Rome 017

Rome 04646-75µm fractionsRome 042Li21-75 and 76-150μm (1 fraction; 3 fractions)

046

017

042

Lumps: 1716 15320-390 (42,6% 1σ)

+/-

Terracina 001Li R_Date(1909,35)95.4% probability

19 (95.4%) 214calAD

200 100 1calBC/1calAD 101 201 301 401

Calibrated date (calBC/calAD)

1600

1800

2000

2200

Rad

ioca

rbon

det

erm

inat

ion

(BP

)

OxCal v4.1.7 Bronk Ramsey (2010); r:5; Atmospheric data from Reimer et al (2009);