Radiocarbon Course Students: These slides are provided for your reference. If you wish to use any of these slides, please make sure to contact the appropriate.

Radiocarbon Course Students:These slides are provided for your reference. If you wish to use any of these slides, please make sure to contact the appropriate lecturer first for permission and attribution. Thank you.

Radiocarbon in Ecology and Earth System Science

W.M. Keck Carbon Cycle Accelerator Mass Spectrometry Facility

Lab Instructors: Guaciara dos Santos-Winston Xiaomei Xu

Lecture Instructors: Sue Trumbore*, Ted Schuur*, John Southon, Ellen Druffel, Jim Randerson, Erv Taylor

Logistics: Morgan Sibley

* Course coordinators

Goals of the class• Learn about the Earth’s carbon cycle from a 14C

perspectiveLectures on what is learned from 14C in Ocean, Atmosphere, Land, Paleo C cycles

• Introduce you to the details of interpreting radiocarbon data

Problem Sets – how does the number you get help answer your question

• Preparation of samples for radiocarbon dating by accelerator mass spectrometry (AMS)

Laboratory methods – taking and analyzing the sample so that you know what you are measuring

Outline of Today’s Lecture

• Global cycles of carbon and 14C

• Isotope basics – how radiocarbon is made and distributed in the environment

• Reporting/Interpreting of radiocarbon data (an intro to problems part of the course)

• Steps involved in making a 14C measurement (a brief introduction to the lab part of the course)

• A brief orientation to the building and AMS lab

Earth System Science is the study of Earth as a coupled and interacting system ofLandAtmosphereHydrosphereBiospherePeople

Involves: physics of transport phase transformations chemical interactions biological reactions and human resource use

Forms of Carbon in the Earth SystemForms of Carbon in the Earth System

AtmosphereAtmosphere (750) Carbon dioxide (gas) CO (750) Carbon dioxide (gas) CO22

(7)(7) Methane (gas) CH4

Ocean (38,000) mostly dissolved ions (HCO3-

(bicarbonate) and CO32- (carbonate)

LandLand (650) Living organic matter (650) Living organic matter (1500) Dead organic matter (soil)(1500) Dead organic matter (soil)

Solid Earth

Land, air, water

Fossil organic matterFossil organic matter (28,000,000) coal, (28,000,000) coal, petroleum, natural gaspetroleum, natural gasLimestone (~50,000,000) CaCO3

ATMOSPHERIC CO2

640 X1015 g C

LIVING BIOMASS

830 X1015 g C

DISSOLVED ORGANICS

1500 X1015 g C

ORGANIC CARBON IN SEDIMENTS AND SOILS

3500 X1015 g C

CO2 DISSOLVED IN OCEANS

38,000 X1015 g C

LIMESTONE AND SEDIMENT CARBONATES

18,000,000 X1015 g C

TRAPPED ORGANIC CARBON: NATURAL GAS, COAL PETROLEUM, BITUMEN, KEROGEN

25,000,000 X1015 g C

Distribution of Carbon;

1015 grams =

1 Petagram (Pg)

Response times are seasons to centuries

Response times are centuries to millennia

Response times are tens of thousands to millions of years

Timescales we can address with 14C

Changes in CO2 on thousand year timescales – glacial to interglacial change indicates past changes in C cycle linked

to climate

Met

han

eT

emp

erat

ure

C

arb

on D

ioxi

de

http://www.realclimate.org/epica.jpg

http://scrippsco2.ucsd.edu/graphics_gallery

0

200

400

600

800

1000

-400000 -300000 -200000 -100000 0

Ca

rbo

n d

iox

ide

(p

art

s p

er

mil

lio

n)

Vostock ice coreTaylor DomeMauna Loa

IPCC high scenario2100, 975 ppm

IPCC low scenario2100, 540 ppm2009, 387 ppm

1959, 316 ppm

Where we are predicted to end up in the next century is far outside the ‘norm’ - what kinds of climate/CO2 feedbacks will operate in future?

Years before present

What makes us sure current CO2 increase is caused by humans? Depletion of Atmospheric 14C – the Suess Effect

Fossil fuel contains zero radiocarbon (millions of years means so many half-lives of 14C that it is all decayed away – so adding CO2 derived from fossil fuel burning reduces 14C over time

Tree rings

Preindustrial atmosphere

Fewer 14C atoms per 12C atom in CO2

Deforestation: Clearing of forests (formerly in the northern hemisphere, now in the tropics)

Responsible for ~40% of total C emissions since 1850

In 1990s 0.5 to 2 GtC/year (8-25% of total emissions)

Source: Ralph Keeling, http://scrippsco2.ucsd.edu/graphics_gallery

Where does the rest go?Also, what happens to CO2 from deforestation (not counted here)?

Carbon dioxide mixing ratio (parts per million)1 ppm = 1 liter CO2 in 1,000,000 liters air

Fossil C Where it goesEmissions

Gig

aton

s of

Car

bon

per

yea

r

Fossil C Where it goesEmissions

Atmo-sphere

OceanAtmo-sphere

Ocean

Land

1989 – 2003/71960 – 1988

Fossil Fuel Emission

Observed atmosphere

increase

Land and Ocean Sinks= +

?

?

Sarmiento et al. 2010

Some of the emitted CO2 is dissolving in the oceans (tomorrow’s lecture)

Surface waters equilibrate quickly; CO2 reacts with water

Falling particles move organic carbon into the deep ocean

Sinking waters in polar regions isolate water that has equilibrated at the surface, removing CO2 for thousands of years

Fossil C Where it goesEmissions

Gig

aton

s of

Car

bon

per

yea

r

Fossil C Where it goesEmissions

Atmo-sphere

OceanAtmo-sphere

Ocean

Land

1989 – 2003/71960 – 1988

Fossil Fuel Emission

Observed atmosphere

increase

Land and Ocean Sinks= +

Sarmiento et al. 2010

Sarmiento et al. 2010 Biogeosciences 7, 2351-2367

Net Land Flux

Deforestation Land

Uptake= -

Year-to-year variation in land uptake Is ± 3 PgC yr-

1

Foster, Motzkin and Slater 1998

Forest Cover in Massachusetts 1830 to 1985

Processes on Land that could be taking up the residual CO2:

- Regrowth of some forests that were previously cut

- Thickening of forests because of forest fire suppression?

- Fertilization of forests by increased N deposition or CO2

- Improved agricultural management

Some big questions for the future

• Can we count on ocean and land sinks to continue to take up ~ half of the CO2 we emit?

• What process(es) is (are) responsible for the land uptake?

• What feedback mechanisms could lead to large changes in the future C cycle?

• What can we learn from past changes in the C cycle?

More details in the week’s lectures…..

Carbon Isotopes – stable and radioactive

The naturally occurring isotope of carbon:

C-12 (98.8%) 6 protons, 6 neutrons

C-13 (1.1%) 6 protons, 7 neutrons

C-14 (< 10-10 %) 6 protons, 8 neutrons

14C is the longest lived radioactive isotope of C, and decays to 14N by emitting a particle (electron):

)(147

146 energyNC

13C – patterns in the environment reflect

mass-dependent fractionation (partitioning among phases at equilibrium and

differences in reaction rates)

14C – Reflects time or mixing Mass-dependent fractionation is

corrected out of reported data using 13C

Isotopes of C contain independent information

Stable C isotope – 13CStable isotope (13C) fractionation:

Kinetic effects: 13C reacts more slowly than 12C 13CO2 diffuses more slowly than 12CO2

Equilibrium effects:12CO2 + 13CO3

2- +H2O = 13CO2 + 12CO32- +H2O

13C will partition into the species where overall energy is lowest (strongest bond or phase with least randomness).Reaction rates and equilibrium partitioning coefficients

are dependent on state variables like T, P

Isotope data are expressed as ratios (e.g. rare isotope/abundant isotope)

compared to a standardIt is nearly impossible to measure the absolute

abundances of isotopes accurately, but differences in relative abundance from one sample to another are easier to measure

For measurements made by different laboratories to be comparable, data are expressed as the ratio to a universally accepted standard

C

C

light

heavy

abundant

rareR

12

13

x10001R

Rx1000

R

RRδ

standard

sample

standard

standardsample

Element Standard R

Carbon Pee Dee Belemnite (calcium carbonate)

13C/12C = 0.011237218O/16O = 0.002671

Carbon-13 nomenclature

By definition, the standards have = 0‰

A leaf with 13C value of –28 ‰ has an “isotope ratio” (Rplant)

PDB

plant

C

C

C

C

12

13

12

13

of (-28/1000) + 1, or 0.972.

Calcium carbonate with 18OPDB of +2 ‰ has

PDBl

shell

O

O

O

O

16

18

16

18

Of (+2/1000) +1

= 1.002

Typical range of 13C values in nature

http://scrippsco2.ucsd.edu/graphics_gallery

Fossil fuel has 13C of -21 to -27 per milIf all emissions are taken up by the biosphere, the 13C of atmospheric CO2 should not change. Dissolution in the ocean preferentially removes 12C more than 13C, so we would expect a decline in 13C of atmospheric CO2

http://scrippso2.ucsd.edu/plots

Decline in O2 is faster than increase in CO2

Stoichiometry says O2/CO2 for fossil fuel burning/biosphere should be ~-1.1

O2 is less soluble than CO2 – so it also provides a way to constrain relative ocean and land sink strengths (i.e. it is like another ‘isotope’ of C )

Unlike stable isotopes, which can be moved around but are always conserved, radiocarbon

is constantly created and destroyed

Total number of 14C atoms (N)

Production in the stratosphere

Loss by radioactive decay

- N

The total amount of radiocarbon on Earth can (and does) vary (Friday’s lecture)

14C is continually produced in the upperatmosphere by nuclear reaction of nitrogen with cosmic

radiation. A smaller amount is produced by cosmic rays interacting with atoms in minerals at the Earth’s surface – we will

ignore that in this class

Cosmicray

spallationproducts

thermalneutron

proton

14Nnucleus

14Cnucleus

Oxidation,mixing

14CO2

stratosphere

troposphere

Ocean/biosphereexchange

Amount of carbon (x1016 moles) 6 1.01.0 1.7-2.0%

typical pre-industrial ratio of typical pre-industrial ratio of 1414C/C/1212C divided by the C divided by the

Modern (i.e. atmospheric) Modern (i.e. atmospheric) 1414C/C/1212C ratioC ratio

per cent of total per cent of total 1414C in the major C in the major global C global C reservoirsreservoirs

Atmosphere (CO2)

280 0.840.84 65-78%Deep Ocean (DIC)

10 0.60.6 2%DOC

30 0.950.95 8-10%Surface Ocean (DIC) 6 0.970.97 1.6-2%

Terrestrial Biota

13 0.900.90 3-4%

Soil Organic Matter

7-70 0.950.95 2-18%Coastal / Marine Sediment

Where the 14C is depends on (1) how much C is there (2) how fast it exchanges with the atmosphere

Radiocarbon is made a second way – from high energy in nuclear

explosions “bomb 14C”

http://www.iup.uni-heidelberg.de/institut/forschung/groups/kk/14co2.html

Radiocarbon is useful on several timescales

Cosmogenic 14C(Radiocarbon dating)

>300 years to ~60,000 years

(± 20-100 years)

Model residence time based on comparison of

14C with Modern C

“Bomb” 14CProduced by atmospheric thermonuclear weapons

testing

~ 1950 to present(± 1-2 years)

Compare 14C to known record of change in

atmosphere

Purposeful tracer 14C follow added radiocarbon

Minutes to years, depending on activity of

tracer

Allows tracing of specific pathways of allocation and

resource use

Radiocarbon data are reported as the ratio of 14C/12C with respect to a standard with known 14C/12C ratio:

95.0

ModernFraction

19- OX1,12

14sample,-25

12

14

C

C

C

C

Ninety-five percent of the activity of Oxalic Acid I from the year 1950

(“Modern” is 1950)

Why does radiocarbon data reported as Fraction Modern or 14C have a correction for mass

dependent fractionation?

Leaf13C = -28 ‰

CO2 in air13C = -8 ‰

14C -12C mass difference is twice that of 13C – 12CTherefore a 20 ‰ difference in 13C

means ~ 40 ‰ difference in 14CExpressed in 14C years this is an apparent age difference of -8033*ln(.96) = 330 years

2

1000δ

1

1000(-25)

1

δ‰sample,C12C14

25‰]sample[C12C14

From Stuiver and Polach 1977

The 14C standard : Oxalic Acid I• The principal modern radiocarbon standard is

N.I.S.T Oxalic Acid I (C2H2O4), made from a crop of 1955 sugar beets.

• Ninety-five percent of the activity of Oxalic Acid I from the year 1950 is equal to the measured activity of the absolute radiocarbon standard which is 1890 wood (chosen to represent the pre-industrial atmospheric 14CO2), corrected for radioactive decay to 1950. This is defined as Modern, which is ~1.12 14C atom for every trillion 12C atoms

• A range of standards with different 14C/12C ratios is maintained by the International Atomic Energy Agency (IAEA).

The ways we use radiocarbon to study the carbon cycle:

Determining the age of C in a closed systemage of pollen, foraminifera, seeds

As a source tracer: mixing of sources with different 14C signatures

For open systems, a measure of the rate of exchange of C with other reservoirs

As a (purposeful) tracertracing pathways (allocation) or rates

We use different ways of expressing radiocarbon data for each of these applications

Different ways to report 14C data depend on the application (Stuiver and Polach 1977)

Expressions that do not depend on the year you make the measurement or take the sample:

Fraction Modern (FM) 0.80

Per cent modern (100*FM) 80%

D = (FM – 1) * 1000 -200 ‰

(this is equivalent to the stable isotope notation)

Radiocarbon age (calculated using FM)

0

0.2

0.4

0.6

0.8

1

0 20000 40000

Radioactivity = number of decays per unit time = dN/dt

dN/dt = -14N,where N is the numberof 14C atoms;

dN/N = -14dt

T = (-1/ 14)ln (N(t)/N(0))

If radiocarbon production rate and its distribution amongatmosphere, ocean and terrestrial reservoirs is constant,Then N(0) = atmospheric 14CO2 value.Note that N(t)/N(0) is then the Fraction Modern (F)[Prove to yourself: 1/2 = ln(2)/14 ]

F

Years1/2

Drops to 0.5 in 5730 years (1/2)

Drops to 0.25 in 2*1/2 years

Radiocarbon Age: used for closed systems in which carbon has resided for hundreds to thousands of years

Radiocarbon Age = -(1/14)*ln(F)

Where F is Fraction Modern and 14 is the decay constant for 14C

The half life (1/2 = ln(2)/14) used to calculate radiocarbon ages is the one first used by Libby (5568 years).

A more recent and accurate determination of the half-life is 5730 years. To convert a radiocarbon age to a calendar age, the tree ring calibration curve is used (we’ll do a problem on this tomorrow).

More on this in the History lecture later …..

Time 1950

Sample made or collected before

1950

14C

/12C

rat

io

0.95*OXI

2007

Fraction Modern, D, 14C age all report the ratio in the year of measurement, which will not vary as time goes on because radiodecay in standard and sample occurs at the same rate ().

Ages are always reported as years before 1950, but the ratio will be the same as that measured in 2007

Time 1950

Sample made or collected in year T

14C

/12C

0.95*OXIin 1950

2008

One of the applications is to figure out past changes in the 14C of atmospheric CO2 using known-age samples; for this we

use the same asC pre-1950, but not after 1950Correct for decay of 14C between T and 1950

decay correction

Sample can be measured any time after 1950, and will have the same ratio as 1950

known-age corrected samples )

1000 1

95.0

exp

19- OX1,

12

14

sample,-25

12

148267

x) - (1950

CC

CC

expresses the radiocarbon signature relative to “Modern” had the sample been measured in 1950. This is useful for studies attempting to show how the radiocarbon signature of air (tree rings) and water (corals) changes with time. It is the basis for creating the calibration curves used to calculate calendar age from radiocarbon age

Corrects for decay of 14C in the sample from the year of growth (x) to 1950)

T)exp(FF

thus

known, are Fand T

lnT

sampleatmosphere

F

F

λ1

atmosphere

sample

sample

Past Changes in Atmospheric 14C recorded in tree rings

T = known age (years before 1950); = ln(2)/5730 yr(actual half-life)

This is equal to 1/8267yr (we refer to this as the mean life)

1950

1950 - T

Decay correction

Past Changes in Atmospheric pre-195014C) recorded in tree rings

Year BP

(a

lso

14C

) of

atm

osp

her

e

Calendar Year

If we know the year the sample was formed, we can correct for radiodecay from that year to 1950 to determine what the 14C of the atmosphere was in the past.

Note that 8000 yrs ago, 14C was about 10% higher than in 1950;

Higher production rates or different distribution of radiocarbon among atmosphere, ocean and land?

Tree-ring calibration curve

The 14C value measured in tree rings of known age is used to determine the 14C value of the atmosphere for the year of tree growth

14C age

Calibration curve for radiocarbon ages shows lack of ability to determine differences in calendar ages using 14C in the past ~300 years.

Radiocarbon age:120 +/-50 years

Yields calibrated ages of 270-160 and 150-50 years BP (Present is always 1950)

For geochemical modeling, especially involving the distribution of bomb 14C, we need a way or reporting the absolute amount of 14C in the sample: Absolute per cent Modern or 14C

-Requires defining a standard that does not change with time: decay-correct the oxalic acid standard to what it would have been in 1950 (i.e. add back in the radiocarbon that decayed in the standard since 1950)

-The value will therefore depend on the year in which the measurement was made (as long as the measurement was made since 1950).

Most commonly used is 14C (geochemical reporting)

Corrects for decay of OX1 standard since 1950Value of this term is 1.0074 in 2011

14C of a sample measured in 2011 will be 7‰ less than if it was measured in 1950 (because 14C in the sample has undergone radioactive decay but the standard has a fixed value).

Difference from is that there is no correction for radiodecay in the sample

Time 1950

14C

/12C

0.95*OXI

Correction for decay of standard since 1950

2009

Standard value doesn’t change with time

14C reports the 14C/12C ratio in the year of measurement compared to the standard measured in 1950. 14C will change for the same sample measured in different years.

Why on earth would you want to use 14C? To perform mass balance – this is called the

‘geochemical’ notation

Total number of 14C atoms in

1963 (bombs)

All produced in atmosphere

Atmosphere

Ocean

Land

Fate of bomb 14C atoms in

2011Radio-decay

Models that trace the fate of bomb

14C require a common

standard that does not change

– those models track radiodecay

directly and therefore can be

directly compared to measurements

1963 2011 Future year

14C decay14C[C1F1 +C2F2 +C3F3]Closed System – Buried CaCO3 crystal

Open System, heterogeneous

Fatm*I FDIC*O

14C decayFDIC*[DIC]*Vol*14C

14C decayCCaCO3*14C

Open System, homogeneous

I*Fleaf

CaCO3 sedimentS*FDIC

Gas Exchange Litterfall Decompositionk1C1F1 +k2C2F2 +K3C3F3

More on Modeling in Thursday’s Lecture…

Closed System – Buried CaCO3 crystal

14C decayCCaCO3*14C

What does 14C tell you?

Complications:

“Hard water effect”

DIC may not be in equilibrium with the atmosphere

i.e. FMDIC ≠ FMatm

Calibrated radiocarbon age can give the time since C in the CaCO3 was buried, assuming the initial FMCaCO3 was zero.

Even if the core top 14C age is not zero, a plot of age versus depth may give sedimentation rate

Fatm*I FDIC*O

14C decayFDIC*[DIC]*Vol*14C

Open System, homogeneous

Gas Exchange

CaCO3 sedimentS*FDIC

Radiocarbon age of DIC gives reflects the relative rates of gas exchange, sedimentation and radio-decay of 14C

14C

Bomb radiocarbon – cannot assume Fatm is constant

14C decay14C[C1F1 +C2F2 +C3F3]Closed System – Buried CaCO3 crystal

Open System, heterogeneous

Fatm*I FDIC*O

14C decayFDIC*[DIC]*Vol*14C

14C decayCCaCO3*14C

Open System, homogeneous

I*Fleaf

CaCO3 sedimentS*FDIC

Gas Exchange Litterfall Decompositionk1C1F1 +k2C2F2 +K3C3F3

More on Modeling in Thursday’s Lecture…

The lab portion of the course

Two ways to measure 14C(1) Beta-decay counting (14C → 14N + -): Measure

radioactivity (decay constant times the number of 14C atoms) directly (compare activity to oxalic acid).

(2) Accelerator mass spectrometry (AMS) Count individual 14C atoms to get 14C/12C ratio. (some

labs measure 14C/13C ratio and use 13C/12C to calculate 14C/12C)

One gram of “Modern" carbon produces about 14 beta-decay events per minute. To measure the age of a 1g sample to a precision of +/- 20 years one needs 160,000 counts, or about 8 days of beta-counting.

AMS allows you to do the same measurement on a 1 milligram sample in a few minutes.

Sample preparation Two ways to measure radiocarbon

Decay Counting Convert C to CO2, then to acetylene (gas) or benzene

(liquid). Requires about 3 grams of sample

AMS Convert C to CO2, then reduce catalytically to

graphite using iron (Fe) catalyst. We use two methods for reduction (zinc vs. H2 as the reductant)

Gas sources (CO2) are starting to be in routine use

How do we measure/ report our Errors? (Accuracy and Precision)

Accurate and Precise

C D

BPrecise but not Accurate

Low precision and low accuracy

Low precision but accurate

• Background – Processing a sample that should contain no radiocarbon – this is a measure of 14C added during processing (graphite production, combustion, etc.).

• Precision – how well do I reproduce the same sample measured more than once? (precision for replicate samples (e.g. soil CO2 sampled in three locations) is likely is not as good as the precision of the AMS measurement (measurement of aliquots of the same CO2)

• Accuracy – how well do I reproduce the known value of a standard material when I run it as an unknown? There are a number of standard materials for purchase from IAEA representing a range of materials and 14C contents.

Factors to consider

What are the stages a sample goes through when it is measured for 14C?

Taking the sample

(Mon/Tues)

What is the question being asked?Does the sample really allow you to answer it?Does the processing in the lab introduce artifacts?

(If needed) Pretreatment

and Combustion

(Mon/Tues)

Purification of CO2 and

conversion to graphite

(Wednesday/Thursday)

Measurement by AMS and

data reduction

(Friday)

Selecting standard

s and blanks to test your sampling procedur

e

(Monday)

Put standard and blank materials through all processes in parallel

(Is my lab 14C clean?)

(Tues-Thursday)

How standards and blanks are used in

data reduction

(Friday)