Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

J. Res. Natl. Inst. Stand. Technol. 109, 185-217 (2004)]

The Remarkable Metrological History of

Radiocarbon Dating [II]

Number 2

Volume 109
Lloyd A. Currie
National Institute of Standards
and Technology,
Gaithersburg, MD 20899-8370
U.S.A.
lloycl.currie@nist.gov

March-April 2004

This article traces the metrological history

of radiocarbon, from the initial breakthrough devised by Libby, to minor (evolutionary) and major (revolutionary)
advances that have brought C measurement from a crude, bulk [8 g carbon] dating
tool, to a refined probe for dating tiny
amounts of precious artifacts, and for
"molecular dating" at the 10 ng to 100 \ig
level. The metrological advances led to
opportunities and surprises, such as the
non-monotonic dendrochronological calibration curve and the "bomb effect," that
gave rise to new multidisciplinary areas of
application, ranging from archaeology and
anthropology to cosmic ray physics to
oceanography to apportionment of anthropogenic pollutants to the reconstruction of
environmental history.
Beyond the specific topic of natural C,
it is hoped that this account may serve as a
metaphor for young scientists, illustrating
that just when a scientific discipline may
appear to be approaching maturity, unanti-

cipated metrological advances in their own

chosen fields, and unanticipated anthropogenic or natural chemical events in the
environment, can spawn new areas of
research having exciting theoretical and
practical implications.

Key words: accelerator mass spectrometry; apportionment of fossil and biomass

carbon; "bomb" C as a global tracer; dual
isotopic authentication; metrological
history; molecular dating; radiocarbon
dating; the Turin Shroud; SRM 1649a.

Accepted: February 11,2004

Available online: http://www.nist.gov/jres

Contents
1.
2.
3.
4.

5.
6.

Introduction
The Birth of Radiocarbon Dating
2.1 Standards and Validation
Natural Variations
The Bomb
4.1 Excess C as a Global Geochemical Tracer
4.2 The Second (Geochemical) Decay Curve
of C: Isotopic-Temporal Authentication
Anthropogenic Variations; "Trees Pollute"
5.1 Fossil-Biomass Carbon Source Apportionment
Accelerator Mass Spectrometry

6.1 The Invention

199
6.2 The Shroud of Turin
200
Emergence of |X-Molar C Metrology
204
7.1 Long-Range Transport of Fossil
and Biomass Aerosol
205
7.2 Isotopic Speciation in Ancient Bones
and Contemporary Particles
210
7.2.1 Urban Dust (SRM 1649a); a Unique
Isotopic-Molecular Reference Material ... .211
Epilogue
214
References
215

186
186
189
190
192
193
193
195
196
199

185

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

1.

Introduction

(111 pages of text) captured the essence of the path to

discovery: from the initial stimulus, to both conceptual
and quantitative scientific hypotheses, to experimental
validation, and finally, to the demonstration of highly
significant applications. The significance of Libby's
discovery, from the perspective of the Nobel
Committee, is indicated in Fig. 1, which includes also a
portrait of Libby in the year his monograph was published [3].' The statement of the Nobel Committee
represents an unusual degree of foresight, in light of
unsuspected scientific and metrological revolutions
that would take place in ensuing years.
Like many of the major advances in science.
Radiocarbon Dating was bom of Scientific Curiosity.
As noted by Libby in his Nobel Lecture, "it had its
origin in a study of the possible effects that cosmic rays
might have on the earth and on the earth's atmosphere"
[4]. Through intensive study of the cosmic ray and
nuclear physics literature, Libby made an important
series of deductions, leading to a quantitative prediction of the natural ''^C concentration in the living biosphere. As reviewed in chapter I of Libby's monograph,
and in the Nobel Lecture, the deductive steps included:
(1) Serge Korff's discovery that cosmic rays generate
on average about 2 secondary neutrons per cm^ of the
earth's surface per second; (2) the inference that the
large majority of the neutrons undergo thermalization
and reaction with atmospheric nitrogen to form ''*C via
the nuclear reaction "*N(n,p)"*C; (3) the proposition that
the ''*C atoms quickly oxidize to ''^COj, and that this
mixes with the total exchangeable reservoir of carbon
in a period short compared to the ca. 8000 year mean
life of ''*C. Based on the observed production rate of
neutrons from cosmic rays (ca. 2 cm"^ s~'), their near
quantitative transformation to "*C, and an estimate of
the global carbon exchangeable reservoir (8.5 g/cm^),
Libby estimated that the steady state radioactivity concentration of exchangeable ''*C would be approximately [(2 X 60)/8.5] or about 14 disintegrations per minute
(dpm) per gram carbon (ca. 230 mBq g"'). Once living
matter is cut off from this steady state, exponential
nuclear decay will dominate, and "absolute dating" will
follow using the observed half-life of ''*C (5568 years).^

This article is about metrology, the science of

measurement. More specifically, it examines the
metrological revolutions, or at least evolutionary milestones that have marked the history of radiocarbon
dating, since its inception some 50 years ago, to the
present. The series of largely or even totally unanticipated developments in the metrology of natural "*C is
detailed in the several sections of this article, together
with examples of the consequent emergence of new
and fundamental applications in a broad range of
disciplines in the physical, social, and biological
sciences.
The possibility of radiocarbon dating would not have
existed, had not '* had the "wrong" half-lifea fact
that delayed its discovery [1]. Following the discovery
of this 5730 year (half-life) radionuclide in laboratory
experiments by Ruben and Kamen, it became clear to
W. F. Libby that "*C should exist in nature, and that it
could serve as a quantitative means for dating artifacts
and events marking the history of civilization. The
search for natural radiocarbon was itself a metrological challenge, for the level in the living biosphere
[ca. 230 Bq/kg] lay far beyond the then current state of
the measurement art. The following section of this
article reviews the underlying concepts and ingenious
experimental approaches devised by Libby and his
students that led to the establishment and validation of
the "absolute" radiocarbon technique.
That was but the beginning, however. Subsequent
metrological and scientific advances have included: a
major improvement in '* decay counting precision
leading to the discovery of natural '* variations; the
global tracer experiment following the "pulse" of
excess "*C from atmospheric nuclear testing; the growing importance of quantifying sources of biomass and
fossil carbonaceous contaminants in the environment;
the revolutionary change from decay counting to atom
counting (AMS: accelerator mass spectrometry) plus
its famous application to artifact dating; and the
demand for and possibility of '* speciation (molecular
dating) of carbonaceous substances in reference
materials, historical artifacts, and in the natural
environment.

2.

Figure 1 shows Libby as the author first met him, shortly after the
latter entered the University of Chicago as a graduate student in
chemistry.
Production rate and reservoir parameters are taken from the Nobel
lecture [4]; these values differ somewhat from those used by Libby
in [5] and in the first edition of his book [2]. The half-life (5568 a) is
the "Libby half-life" which by convention is used to calculate "radiocarbon ages;" the current accepted value for the physical half-life is
(5730 40) a [5a].

The Birth of Radiocarbon Dating

The year before last marked the 50th anniversary of

the first edition of Willard F. Libby's monograph.
Radiocarbon Dating^published in 1952 [2]. Eight
years later Libby was awarded the Nobel Prize in
Chemistry. In a very special sense that small volume
186

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

"Seldom has a single

discovery in chemistry had
such an impact on the
thinking in so many fields
of human endeavor."
-Nobel Committee (1960)
W. F. Libby(ca. 1952)
Fig. 1. Portrait of W. F. Libby, about the time of publication of the first edition of his monograph, Radiocarbon Dating
(1952), and statement of the Nobel Committee (1960) [3].

thermal diffusion enrichment technique [6] was not: it

demanded very large samples and thousands of (1946)
US dollars "to measure the age of a single mummy"
[4]. Development of an acceptable technique was
formidable, as outlined in Table 1. A substantial increase in signal was achieved by converting the sample
to solid carbon, which coated the inner wall of a
specially designed "screen wall counter;" but the background/signal ratio (16:1) still eliminated the possibility of meaningfiil measurements. At this point, Libby
had an inspiration, from the analysis of the nature of the
background radiation [4]. He concluded that it was
primarily due to secondary, ionizing cosmic radiation
having great penetrating powernegative mu mesons
{\r). By surrounding the sample counter with cosmic
ray guard counters operating in an anti-coincidence
mode, most of the \r counts could be eliminated, resulting in a further background reduction by a factor of
twenty, to approximately 5 counts per minute (cpm).
The final background to signal ratio of 0.8 for living
carbon, made possible the measurement of natural
(biospheric) "*C with a precision under 2 % (Poisson
relative standard deviation) with a total (sample, background) counting time of just 2 d ([2], Chap. V). Fig. 3
shows the low-level counting apparatus devised by
Libby, with which the seminal ''*C dating measurements

Two critical assumptions are needed for absolute '*

dating: constancy of both the cosmic ray intensity and
size of the exchangeable reservoir on average for many
thousands of years. A graphical summary of the above
points is given in Fig. 2.
Libby first postulated the existence of natural '* in
1946, at a level of 0.2 to 2 Bq/mol carbon (1 dpm/g to
10 dpm/g) [5]. His first experimental task was to
demonstrate this presence of "natural" ''^C in living
matter. The problem was that, even at 10 dpm/g, the '*
would be unmeasurable! The plan was to search for
natural '* in bio-methane, but the background of his
well-shielded 1.9 L Geiger counter (342 counts per
minute) exceeded the expected signal by a factor of
400. Libby and coworkers did succeed in demonstrating the presence of ''^C in living matter, however. For an
account of their creative approach to the problem, see
their one page article in Science, "Radiocarbon from
Cosmic Radiation" [6].^
Having detected '* in the living biosphere, Libby
and his colleagues had to develop a measurement
technique that was both quantitative and practical. The

To fully appreciate the nature of the experimental impediments and

flashes of insight along the path to discovery, students are encouraged to study the original scientific literature, as given here, rather
than restricting attention to subsequent summaries in textbooks.

187

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

PRODUCTION OF ^^C
p
R
O
D
U
C
T
I
O
N

Cosmic
Ray

D
I
S
T
R
I
B
U
T
I
O
N

*C

D
E
C
A
Y

19

Equilibrium Concentration: T^ = 10 "^

14K

Tiien: ^"^0

*N + e' + V

T^^2 - 5700 years

One Gram - -10 counts/minute

Fig. 2. Graphical representation of the production, distribution, and decay of natural

(courtesy of D. J. Donahue).(Parameter values are approximate.)

were made. The "*C screen wall counter is visible

through the open, 8 inch thick cantilevered steel doors
having a wedge-like closure. The steel "tomb" reduces
the background by about a factor of five. The bundle
of anticoincidence cosmic ray guard counters, seen
surrounding the central counter in the figure, eliminates
some 95 % of the residual background from the
penetrating \i radiation, through electronic cancellation.

Table 1. Libby's Measurement Challenge

Cosmic ray neutron intensity: 2 n cm s

Exchangeable carbon reservoir: 8.5 g cm
Estimated C activity: 14 dpm g~ (0.23 Bq g~ )
Sample size (detector efficiency): 8 g carbon (5.5 %)
Estimated modem carbon rate 6.2 cpm (min~ )
Background rate: 500 cpm (unshielded), 100 cpm (20 cm Fe)

Assumptions:
Constant production rate
Fixed exchangeable C reservoir (uniform distribution)

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

step was to measure the "*C concentrations in selected

historical artifacts of known age, and compare them to
the "absolute" "*C age. The latter was accomplished by
comparing the artifact "*C concentration (dpm/g C) to
that of the living biosphere. The absolute age derives
from the inversion of first order nuclear decay relation,
using 15.3 dpm/g and 5568 a as the parameters of the
"absolute" natural '* decay curve.
The famous result, utilizing known age tree rings and
independently-dated Egyptian artifacts, is shown in
Chapter I of Libby's 1952 monograph and Fig. 4 in this
article. Although the relative measurement uncertainties are moderately large (ca. 1 % to 5 %), the data
provide a striking validation for the radiocarbon dating
method over a period of nearly 5000 years. Note that
the curve shown is not fit to the data\ Rather, it represents the absolute, two-parameter nuclear decay function. (See [8] for detailed information on the validation
samples selected.)
This initial absolute dating fiinction served to establish the method, but it indicated the need for a universal radiocarbon dating standard, since the reference
value for the intercept (here 15.3 dpm/g) would vary

Fig. 3. Low-level anticoincidence counting apparatus devised by

Libby for the original C measurements that led to the establishment
of the radiocarbon dating technique (Ref [2], and Radiocarbon
Dating (jacket cover) R. Berger and H. Suess, eds., Univ. California
Press, Berkeley (1979).)

Perhaps the most valuable metrological lesson from

Libby's early work was the extreme importance of
formulating a realistic theoretical estimate for the
sought-after "signal." Without that as a guideline for
designing a measurement process with adequate detection or quantification capabilities, there is essentially
no possibility that natural radiocarbon could have
been found by chance with the then current radiation
instrumentation.
2.1

SAMPLES OF KNOWN AGE

TREE RING (SBO A.D.)

13-

Standards and Validation

Once the measurement of natural "*C became

feasible, the immediate task tackled by Libby and his
colleagues was to test the validity of the radiocarbon
dating model. The first step consisted of determining the
zero point of the natural radiocarbon decay curve i.e.,
the radioactivity concentration (dpm "*C per gram C) in
living matter, and to test for significant geographic variation. This was a major component of the PhD thesis of
E. C. Anderson [7]; the result (i?) was (15.3 0.5) dpm/g
[255 Bq/kg] with no significant deviation from the
hypothesis of a uniform global distribution.'* The next

PTOLEMY (200 150 B.C)

TAYINAT(67550B.C)
Jj REDWOOD (979 52 B.C.)

SESOSTRIS (1800 6)
ZOSER (2700175ac)
CURVE CALCULATED
FHOM PRESENT DAY POiNT
AND HALF LIFE OF
RADIOCARBON 5566 30 YEARS

1000

The neutron intensity in the atmosphere, and hence the C production profile, has major variations vertically (because of cosmic ray
absorption with atmospheric depth) and latitudinally (because of
geomagnetic shielding)See Figs. 2 and 3 in Ref [2], Because C
has such a long mean life (=8000 a), however, it was expected that
any residual gradients in the global exchange reservoir would be
undetectable, given the 3 % to 5 % uncertainties of Libby's original
measurements (Ref [2], Chap. I).

SNEFERU(262575B.c)

2000
3000
4000
5000
HISTORICAL AGE (YEARS)

6000

7000

Fig. 4. Radiocarbon dating validation curve (1952): the "curve of

knowns" that first demonstrated that absolute radiocarbon dating
"worked." The validation points represent tree rings and historical
artifacts of known age. The exponential function is not fit to the data,
but derived from the independently measured half-life and the C
content of living matter ([2], Fig. 1).

189

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

among laboratories, if they each made their own

standards. The problem was tackled by the international radiocarbon community in the late 1950s, in cooperation with the U.S. National Bureau of Standards. A
large quantity of contemporary oxalic acid di-hydrate
was prepared as NBS Standard Reference Material
(SRM) 4990B. Its "*C concentration was ca. 5 % above
what was believed to be the natural level, so the
standard for radiocarbon dating was defined as 0.95
times the '* concentration of this material, adjusted to
a "C reference value of-19 per mil (PDB). This value
is defined as "modem carbon" referenced to AD 1950.
Radiocarbon measurements are compared to this
modem carbon value, and expressed as "fraction of
modem" (Z^); and "radiocarbon ages" are calculated
from fi^ using the exponential decay relation and the
"Libby half-life" 5568 a. The ages are expressed in
years before present (BP) where "present" is defined as
AD 1950. A published estimate for the ''^C concentration of "modem carbon" is given as (13.53 0.07)
dpm/g [9]. In July 1983, a replacement SRM 4990C
was substituted for the nearly exhausted SRM 4990B.
It was prepared from oxalic acid derived from the
fermentation of French beet molasses from harvests of
1977. A copy of the Certificate Analysis of SRM
4990C, together with pertinent references, may be
obtained from the website: http://nist.gov/srm [10].'
Libby's successflil development of the science of
radiocarbon dating led to the rapid establishment of
more than a hundred dating laboratories world-wide,
the initiation of a joumal supplement that later became
the joumal Radiocarbon, and the establishment of a
continuing series of triennial RADIOCARBON conferences, the first of which took place in Andover,
Massachusetts in 1954.

3.

This "failure" resulted from basic advances in "*C

metrology. New approaches to low-level counting
yielded measurement imprecision that ultimately
approached 0.2 % (rsd);* and construction of the
"radiocarbon dating calibration curve" from meticulously counted annual tree ring segments showed that
assumptions of constancy within different geochemical
compartments of the exchangeable carbon reservoir,
and over time, were invalid. (This is a classic example
demonstrating that one cannot prove the "null hypothesis;" the validation curve that established the
radiocarbon dating method demonstrated consistency
(validity) only within the errors (uncertainties) of the
validation measurements.) The failure of the absolute
dating model was, in fact, a notable success. The revolutionary discovery of natural radiocarbon variations
literally arose out of the "noise" of absolute radiocarbon dating, and it transformed the study of natural "*C
into a multidisciplinary science, giving rise to totally
new scientific disciplines of "*C solar and geophysics.
At his opening address at the 12th Nobel Symposium
on Radiocarbon Variations and Absolute Chronology
[12] in Uppsala, Nobelist Kai Siegbahn emphasized
that "This subject is [now] interesting to specialists
in many different fields, as can be seen from the list
of participants, showing archaeologists, chemists,
dendrochronologists, geophysicists, varved-clay geologists, and physicists" (Ref. [12], pp. 19f). An early
version of the dendrochronological "*C calibration
curve, presented by Michael and Ralph at the
Symposium, is given in Fig. 5 (Ref [12], p. 110).' The
Bristlecone pine, as shown in the figure, has made a
seminal contribution to the science of dendrochronology, and through that, to the study of natural ''*C variations. It is considered by some to be the world's "oldest
living thing," with a single tree containing annual rings
going back 4000 years or more. It is clear from Fig. 5

Natural Variations

Already, by the time the Nobel Prize was awarded.

Radiocarbon Dating appeared to be approaching maturity, with a rich fiiture in application as opposed to new
fundamental discovery. This all changed, however,
when some of the fundamental assumptions proved to
be invalidwhat might be considered as the "failure of
Radiocarbon Dating."

The deciding factor for high precision C measurement was the

successfiil development of CO2 gas proportional counting, after
several failed attempts. Compared to Libby's solid sample (graphite)
technique, the CO2 method resulted in smaller sample sizes and
efficiency enhancement by nearly a factor of twenty.
The relatively imprecise dendro-calibration curve in Fig. 5 extends
to ca. 5000 BC. Meanwhile, the radiocarbon dating calibration function has undergone considerable refinement: it now comprises an
extensive database, and it has become an essential element of all
radiocarbon dating. The 1986 Calibration Issue of the joumal
Radiocarbon [13] has a compilation going back to ca. 8000 BC.
More recent attempts at extending the record much further back in
time have utilized C comparisons with other dating methods,
notably U/Th disequilibrium dating. By this means, calibration data
have been given for periods beyond 20 000 BC [14].

Several secondary standards for C dating are available through

the International Atomic Energy Agency. These materials, designated
IAEA Cl -C8, consist of wood, cellulose, sucrose, and carbonate;
they cover a range of 0.00 pMC to 150.6 pMC, and have been subject to an international comparison [11]. Note that pMC (percent
modem carbon) refers to_4i expressed as a percentage.

190

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

5600 BC r

4800 BC

4000 BC

111

3200 BC

^
n

2400BC

1600BC

CD

01
800BC

AD/BC

800

AD1600 -.
o
o

9
<

o
o

(j

(J

rfi

iTl

[Q

CD

ro

iTl

rTi

O
CO

CJ
IN
CO

O
O
fl-

O
00
TT

<

y?

o
o
o

Dendro -Age
Fig. 5. Radiocarbon Variations, discovered by comparison of high precision radiocarbon "dates"
with high (annual) accuracy tree ring dates. The plot, which covers the period from about
5000 BC to the present, represents an early version of the radiocarbon dating calibration curve
([12], p.UO). The photo shows the Bristlecone pine, the major source of dendrodates extending
back many millennia (Photo is courtesy of D. J. Donahue).

that the dendrochronological age shows a significant

departure fi-om the absolute '* (nuclear) age, beginning about three thousand years ago, and continuing
through the end of this series of measurements (ca.
5000 BC). These newly discovered deviations from the
absolute dating model, of course, posed new scientific
questions: what are the causes of the deviations, and
can we use them to better understand Nature? In fact,
the dendro-calibration curve serves dual purposes. For
more classic "dating" disciplines, such as archaeology,
anthropology, and geology (event dating), it gives an
empirical correction fiinction for the simple radiocarbon ages (BP) derived from the first order decay
relation. For solar and geophysics and related disciplines, it gives the potential for the quantitative investigation of the causes of the variations.

The Nobel Symposium serves as a rich resource for

information about the natural ''^C variations. An excellent exposition of the three prime causative factors is
given by Hans Suess (Ref [12], pp. 595-605). These
are: "(1) changes in the '* production rate due to
changes in the intensity of the [earth's] geomagnetic
field; (2)... modulation of the cosmic-ray flux by solar
activity; (3) changes in the geochemical radiocarbon
reservoirs and rates of carbon transfer between them."
The major departure (ca. 10 %) seen in Fig. 5 is considered to be due to the geomagnetic field, corresponding
to a factor of two change in its intensity over the past
8000 years [15]. This has given major impetus to the
science of archaeomagnetism. The other two factors
are considered responsible for the partly periodic
fine structure exhibited in the curve, with varying
191

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

RADIOCARBON AND SOLAR ACTIVITY

lOM

1100

1150

1200

^1iO

1300

i:9S0

1400

14S0

ISOO

1550

1600

1650

1?0C

17S0

ISQO

It60

190O

Air temperatures, eastern Europe

1,000 800
YEARS AGO

600

400

200

TODAY

CLIMATE
Fig. 6. Radiocarbon Variations and Climate: the influence of solar activity (sunspot record) (top) on
rates) and climate (Maunder Minimum temperature record) (bottom) [15, 16].

amplitudes of about 1 % to 2 %. (See Figs. 1, 2 in the

Suess article, respectively, for plots of the first order
(geomagnetic) and second order (fine structure) deviations from the ideal exponential decay function ("radiocarbon age")-)
A fascinating link exists between dendrochronology
and radiocarbon age, related to climate. That is, tree
rings by their width time series, like ice cores by their
'*0 time series, give insight into ancient climate [16].
This, in turn, may be linked to the aforementioned '*
variations from changing solar activity and/or variations in geochemical reservoirs. Fig. 6 represents a
famous example of the inter-relationships among solar
activity (sunspots), natural radiocarbon variations, and
climate (Ref. [15], Fig. 5a; Ref. [16], p. 615). The upper
part of the figure shows the correlation between the
sunspot record (circles, and ca. 11 year cycles) and the
"*C variations. The period of low solar activity, and
correspondingly increased "*C activity, peaking at about
1500 AD and 1700 AD is striking. The lower part of

C concentrations (cosmic ray production

the figure suggests a strong link to global climate,

represented here by the "little ice age."

4.

The Bomb

Atmospheric nuclear testing had an unintended but

profound impact on "*C geoscience. It approximately
doubled the "*C concentration in atmospheric COj, and
consequently in living matter, by the mid-1960s. This
came about because neutrons released from nuclear
fission (or fusion) react with atmospheric nitrogen by
exactly the same reaction, '^N(n,p)''*C, as the secondary
neutrons from cosmic rays. The "bomb pulse" of excess
"*C was recorded in all parts of the living biosphere,
from vintage wine [17] to contemporary tree rings [18].
It was characterized by a sharp injection of '* in the
early 1960s, followed by relatively slow geochemical
decay after the limited (atmospheric) nuclear test ban
treaty. Totally new and unanticipated opportunities to
perform global tracer experiments resulted from this

192

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

900
800
700
600
500
400
300
200
100
0
-100
1950

1960

1970

1980

1990

YEAR
Fig. 7. Input function of excess ("bomb") C: a global tracer for carbon cycle dynamics
in the atmosphere, biosphere, and oceans [19].

sudden, widespread injection of anthropogenic '* into

the biogeochemical system.

brings regarding the effects of the oceans on pollutant

and heat transport and climate [22].*

4.1

4.2

Excess "C as a Global Geochemical Tracer

An extensive world-wide program of monitoring the

excess atmospheric ''^COj began with the onset of
nuclear testing and continues today. Results of precise
measurements of the input function for excess ''^COj
are shown in Fig. 7 (Ref[19]; Ref [20], Chap. 31,
(I. Levin, et al.)). Use of this known pulse of excess '*
as a tracer has allowed scientists to study exchange and
transport processes in the atmosphere, the biosphere,
and the oceans on a scale that would otherwise have
been nearly impossible. Simple visual examination of
Fig. 7 shows, for example, that the excess atmospheric
''^C injected in the northern hemisphere gave an attenuated signal in the southern hemisphere, and that there
was a lag time of approximately 2 years.
Nowhere has the bomb pulse been more important
than in furthering our understanding of the dynamics
of the ocean. A comprehensive program (GEOSECS:
Geochemical Ocean Section Study) to follow the plume
of excess '* as it diffused in the Atlantic and Pacific
oceans was initiated in the 1970s. A small example of
the findings is given in Fig. 8, where we find a nearly
uniform distribution below the mixed layer, indicating
rapid vertical transport in the North Atlantic, in contrast
to model predictions [19, 21]. The scientific impact of
this massive tracer study of ocean circulation is striking, considering, for example, the new knowledge it

The Second (Geochemical) Decay Curve of

"C: Isotopic-Temporal Authentication

Geochemical relaxation of the excess atmospheric

"*C after about 1970 has resulted in a second (shortlived) "decay curve" for "*C (tail of the input function.
Fig. 7). This has made possible a new kind of radiocarbon dating, where modem artifacts and forgeries,
food products, forensic biology samples, and industrial
bio-feedstocks can be dated with near annual resolution
[24]. As a result of the new submilligram measurement
capability (Sec. 6), short-term radiocarbon dating is
beginning to achieve commercial importance, as
exemplified by its application to the dual isotopic
("C, "*C) fingerprinting and time stamping of industrial
materials.
A case in point is the Cooperative Research and
Development project between the NIST Chemical
Science and Technology Laboratory and the DuPont
Central Research and Development Laboratory [25].
The advent of accelerator mass spectrometry, as discussed in
Sec. 6 of this article, has given a major boost to our knowledge of
ocean circulation. Information gained through the GEOSECS
program has been greatly amplified in the World Ocean Circulation
Experiment (WOCE), where requisite sample sizes were reduced
from 200 L of sea water each, to less than 1 L; and the C ocean
circulation database grew by more than 10 000 dates during the
1990s [23].

193

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

A^H(%)
-goo
I

-100

-100
1

100

200

prS-EMHTlt)

1972
500

1000

1500

2000

Z50D

3000

3500

model

OEOSECS 27
N. AlloMic

4000

4500

-j

I/w

ll

I-

Fig. 8. Excess C and ocean circulation (GEOSECS). Model (left) and experimental
(right) vertical transects of bomb C in the North Atlantic [19].

fossil feedstock (petroleum) based. The ability to establish a unique isotopic fingerprint for the DuPont
biotechnology materials was critical for the identification of the product as a unique composition of matter,
and to track it in commerce. The work represents a
frontier of high accuracy, dual isotope metrology, with
'^C data Mr<0.01%) serving to discriminate among
different photosynthetic cycles, and '* data {u, < 0.5 %)
serving both for quantitative fossil-biomass apportionment and for dating the year of growth of the biomass
feedstock.
A graphical summary of the results of the project is
presented in Fig. 10, which shows the dual isotopic
signatures of the copolymer (3GT) and bio-sourced

The goal of the project was to demonstrate the capability to authenticate and date renewable (biosourced)
feedstocks, chemical intermediates, and finished
industrial products using high accuracy dual isotopic
("C-"'C) "fingerprinting," traceable to NIST. The
specific project, as outlined in Fig. 9, was directed
toward the unambiguous identification of the copolymer polypropylene terephthalate (3GT)) produced from
the biosourced monomer 1,3-propanediol (3G), which
was derived from com as feedstock. (Terephthalic acid
(TPA) served as the complementary monomer.)
Isotopic discrimination was essential because it is not
possible chemically to distinguish the biosourced 3G
and 3GT from existing industrial materials that are

194

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

Corn

Petroleum

Glucose

Xylene

1, 3-Propanediol

Terephthalic Acid

HOCH2CH2CH2OH

HOOC \^ COOH

Biomass Carbon

Fossil Carbon

[3G]

[TPA]

i
i

1
1

Polymerization
O
O
II /=x
II
-t C -^^-C OCH2CH2CH2O^
[3GT]

Polypropylene Terephthalate
Fig. 9. Polypropylene Terephthalate: biomass and fossil feedstocks. The 1,3, propanediol monomer is derived from a renewable (biomass) feedstock via laboratory biotechnology: conversion of glucose or comstarch using a single microorganism. The copolymer has potential large volume demand, and is useful as a fiber, film, particle, and a
molded article [25].

monomer (3G); as well as values for isotopic reference

materials (SI: SRM 4990B [oxalic acid]; S2: IAEA C6
[ANU sucrose]; S3: SRM 1649a [urban dust])., and
pre-existing materials (3G', 3G"). The dashed line
joining the copolymer end members (3G, TPA) demonstrates isotopic-stoichiometric mass balance. Rectangular regions in red define the "scope of claims"
(authentication regions) for the new isotopic compositions. The blue "x" in the figure represents data for an
independent batch of the monomersent to NIST
"blind" to test the validity of the authentication region
for bio-sourced 3G. The results show both that the
test was successful and that the separate production
batches of the 3G monomer had unique isotopic signatures. The approximately ten-fold expansion of the
isotopic data for two independent batches (A, B) of

corn-glucose (bottom right) demonstrates the dual

isotopic discrimination capability of the technique. In
fact, using the short term "decay" curve of "*C (Fig. 11),
it was possible to date the two batches to the nearest
year of growth, 1994 (A) and 1996 (B), respectively.
(Standard uncertainty bars shown.)

5.

Anthropogenic Variations; "Trees

Pollute"

The achievement of high precision, low background

counting, discussed in Sec. 3, led also to the first
isotopic evidence of global pollution with fossil COj
named the "Suess Effect," after its discoverer. A
dramatic monotonic drop in the ''^C/'^C ratio in tree
rings beginning in the late 19th century, reflecting the
195

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

1.7
S2_

1.4 3G
-

Glucxjse

biomass-C

1.1 -

S1
3G'

^ 0,8
O

--.'
{b)

(c)

--

(a)

/-"

S3

0,5 -_

.--'

3GT j.^'' 3 (d)

0.2 -

3G"

TPA
-0.1

-29

-26

(e)

fossil - C

-23

-17

-20

-14

-11

1.20

0.33

[A]

0.31 [B]

0,29 Glucose (a)

0.27
-25.5

-24.5

10.5 -10.1

-23.5

-9.7

-9.3

:13
(3'X(permil)
Fig. 10. Unique Isotopic Signatures: the C- C plane [25]. The main panel shows dual isotopic signatures f:or (1) NIST (SI, S3) and IAEA (S2)
traceability standards, and (2) glucose from biomass (a), the new bio-sourced monomer 3G (b) (from cornstarch), the resulting copolymer 3GT
(d), and pre-existing products 3G', 30" (c, e). Expanded views of the authentication regions (red rectangles) for the copolymer (left) and monomer
(center) are given in the bottom panels, plus = 10-fold expansion (right) of the isotopic data for independent batches (A, B) of a biomass feedstock
(glucose from com). The blue "x" represents a blind (3G) validation sample.

5.1

use of coal during the Industrial Revolution, showed a

2.5 % fossil carbon dilution effect by the 1950s (Ref
[12], p. 289), after which it was eclipsed by the vast
injection of "bomb" carbon. Thus began still another
field of'* science: the investigation of anthropogenic
variations, particularly as related to environmental
pollution.

Fossil-Biomass Carbon Source Apportionment

Research on more specific local or even regional

carbonaceous pollution began slowly, because of the
massive samples required. Heroic sampling efforts in
the late 1950s demonstrated the principle by measurements of particulate carbon pollution in U.S. urban
atmospheres [26, 27]. After a lapse of two decades.

196

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

1.6
1.5
1.4

^\
\
~

1,2

LI

1.0

\l
VK^^^
^^^^'''^^-^
^'^'^*^""~^--.

1970

19SU

199(1

2000

Fig. 11. Short term C "decay" curve, representing geochemical relaxation of excess
atmospheric C from nuclear testing [Levin et al., in (Ref [19]; Ref [20], Chap. 31).
Information critical for the discussion in Sec. 7.2.1 is indicated by the arrownamely, the
sampling date and corresponding biomass C enrichment for SRM 1649a (urban dust).

research in this area was renewed by the author, stimulated by a 1975 article in Science reporting that the
culprit for a severe case of urban pollution in tidewater
Virginia might be hydrocarbon emissions from trees
[28]. The evidence was chemical and controvertible:
plausible, but circumstantial evidence suggested that
the air pollution was due to hydrocarbon emissions
from trees rather from automobile exhaust or evaporation from nearby industrial and military storage tanks.
The article concluded that "the relatively unsophisticated monitoring of [organic] pollutant concentrations ... will rarely be of value in identifying [pollutant]
sources ..." Recognizing immediately that '* could
function as an undisputed discriminator, we decided to
design miniature low level counters, capable of measuring just 10 mg carbon samples, more than two orders
of magnitude smaller than those used in the two earlier
studies. Apart from forest fires, we found that the trees
were not the prime culprits, except for the case where
humans were using the trees for fuel! A review of
research in this area in the ensuing 20 years is given in
Ref [29].
One illustration of '* aerosol science is given in
Fig. 12. It is drawn from perhaps the most extensive
study to date of urban particulate pollution using "*C.
The multi-year, multidisciplinary study of the origins of
mutagenic aerosols in the atmospheres of several
U.S. cities, focussed on Albuquerque, New Mexico
during the winter of 1984-1985. The photos show the

tremendous impact on visibility from particulate pollution from rush hour traffic. Results of the two month
study of particulate carbon proved that daytime pollution (up to ~ 65 %) was dominated by motor vehicle
emissions (fossil carbon), and nighttime pollution
(up to ~ 95 %), by residential woodbuming (biomass
carbon), with the mutagenicity (potency) of the motor
vehicle particles more severe by a factor of three [30].
Particulate carbon aerosols are now widely recognized
as an extreme health hazard in a number of U.S. cities;
and except for periods dominated by wildfires, major
studies including '* measurements have produced
incontrovertible evidence that the urban episodes are
dominated by fossil carbon, largely from motor vehicle
exhaust [31].
Quantitative apportionment of natural and anthropogenic sources of particulate carbon, methane, carbon
monoxide, and volatile organic ozone precursors in
the atmosphere, meanwhile, has seen a significant
expansion thanks to the sensitivity enhancement
of accelerator mass spectrometry (AMS) [32, 33].
Most recently, with the emergence of micromolar ''^C
AMS, and GC/AMS, the ability to "date" individual
chemical fractions in small samples is having important
impacts on both artifact age accuracy, and our understanding of perturbations of the human and natural
environments by fossil and biomass carbonaceous
species. (See Section 7).

197

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

13
Fig. 12. Anthropogenic C variations: fossil-biomass carbon apportionment of particulate air pollution in Albuquerque, New Mexico. (Photos showing visibility reduction in
early morning (top) and mid-afternoon (bottom) are courtesy of R .K. Stevens [30].)- C
measurements quantified atmospheric soot from motor vehicles and residential woodburning, and helped apportion concomitant data on particulate mutagenicity.

198

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

6.

Accelerator Mass Spectrometry

6.1

The Invention

energy (megavolt) nuclear accelerators were used as

atomic ion mass spectrometers [34-36]. Two measurement ideas held the key: (1) Negative carbon ions are
produced by a sputter ion source, using graphite as the
target. (2) Following low energy mass selection, atomic and molecular negative ions are injected into an
accelerator tube with a megavolt potential. The major
isobar is eliminated because nitrogen does not form a
stable negative ion. Passage of the high energy ions
through a stripper gas or foil destroys all molecular ions
through the "coulomb explosion," leaving only atomic
carbon ions in the +3 or +4 charge state. ''*C/'^C ratio
measurements down to ca. 10"'^ are thus made possible.
Typical sample sizes are 0.5 mg to 1 mg; modem
carbon yields 10 000 counts in just a few minutes;
and instmment backgrounds are negligible (<0.2 %
modem, equivalent to a '* age of >50 000 years BP).
A diagram of the accelerator at one of the leading
facilities is given in Fig. 13 [37]. The dramatic impact
of high energy (atomic ion) mass spectrometry is
shown in Fig. 14, where it is clear that natural "*C is
quite unmeasurable by low energy (conventional) mass
spectrometry due to molecular ions exceeding the
"*C signal by more than eight orders of magnitude
(Ref. [20], Chap. 16]! Excellent reviews of the history,
principles, and applications of AMS are given in
Ref [20] by H. Gove (Chap. 15) and R. Beukens
(Chap. 16).

The second revolution in '* measurement science

was the discovery of a means to count "*C atoms, as
opposed to '* decays (beta particles). The potential
impact on sensitivity was early recognized: inverting
the first order nuclear decay relation, one finds that the
ratio of the number of '* atoms to the number of '*
decays for any given sample is simply (T/t), where ris
the mean life (8270 a for '*), and t is the counting time
used for measurement of the disintegrations. Allowing
for the difference in relative detection efficiency
between AMS and low-level counting, and setting i to 2 d,
gives a sensitivity enhancement of roughly lO'', in favor
of AMS. This implies a dating capability of submilligram amounts of modem carbon.
The prize of radiocarbon dating at the milligram
level was so great that major efforts were made
to refine mass spectrometric techniques to render the
1.2 X 10"'^ ''*C/'^C ratio of modem carbon measurable;
but, like Libby's initial attempt to count natural radiocarbon (without enrichment), natural '* proved
unmeasurable by conventional mass spectrometry.
Impediments Irom molecular ions and the extremely
close isobar (''*N: Am/m = 1.2x10"^) were overwhelming. Success came in 1977, however, when high

heCATIVC ION
SWTTEft SOURCE

MAGNETtC MASS
ANALYZER
ruiituci

EN-TANDEM ACCELERATOR
STHlPPtft

101 ^ \Q\

IIIIIIIIIIIINIIirT^^IIIIIIIIKIIIMMI

Jl^^^iiiiiilllllllllll

MAGNETIC ^4ASS
ANALYZER

EICTROSTAT;C
DEFLECTOR 15*

ELtCTOOSTAnC
00 LEI

CDC3

Nf

OO

IMJUX

surs

_/ UASSC.Il
1,14

/ - VMLK W

M:
I ACE

H CURRENT/FREQ. h
OUTPUT

'c/c

COMPUTER

-|CUBREHT/FIJEO~t- "^'^"^"
"C evENTS

"c/'h

Fig. 13. AMS: tandem accelerator at ETH, Zurich. Negative carbon ions, produced with a Cs sputter ion source, undergo low energy mass
resolution and then are injected into the 4.5 MV accelerator tube. Molecular ions are destroyed by the stripper gas, and emerging 18 MeV C^
beams of C, C, and C are mass analyzed and measured in current (stable C ions) and event ( C ions) detectors [37].

199

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

MASS NUMBER
1

V
K>"'
Kf'

13

-r

X^.Hl
lo"

-f
/'

J
^

xlO

f-CH

3n 10"

\ r
V^ V

sr

10-'

a.

-1

'CH- SxlO""
^CH,- 3x10'" -

"CH;
'"NH-

V,-J

L__yv.

Conventional M

i 10-'
<J
z
o

-8
'0
^9
lo'
10
10

-n
to

W _
-B
10

10^"

CONTEMPORARY
'"C
19000 YEARS

,
019

__^ n
*"
j

0.20

INJECTOR MAGNETIC FIELD

ij1

Accelerator MS
j1

QJI

(TESLA)

Fig. 14. Conventional (top) vs accelerator (high energy) (bottom) mass

spectrometry: C/ C sensitivity is enhanced by more than eight orders
of magnitude through destruction of molecular ions (and unstable N~)
(Ref [20], Chap. 16).

As noted in the reviews by Gove and Beukens, the

AMS revolution has extended well beyond '*, spawning a totally new research area in long-lived isotopic
and ultra trace stable cosmo- and geo-chemistry and
physics through its capability to measure 'H, '*, ^'Al,
^'Cl, '*'Ca, and \, and most recently, selected actinides.
Within one year of the publications announcing successful "*C AMS, another continuing series of international conferences was bom. The first international
AMS conference took place in 1978 in Rochester, New
York. These conferences have continued on a triennial
basis, with each proceedings occupying a special AMS
conference issue of the journal, Nuclear Instruments
and Methods in Physics Research.
6.2

widely accepted statements (1) concerning scientific

investigations of the Shroud, and (2) following publication of the Nature article announcing radiocarbon
dating results (Fig. 15; Ref [38]).
1: "The Shroud of Turin is the single, most studied
artifact in human history."
14

2: "The Nature ( C) article has had more impact on

Shroud research than any other paper ever written on
the subject."
The article, which was prepared by three of the most
prestigious AMS laboratories, is available to the general public on the web {www.shroud.com/nature.htm).
Together with public television [39], it is helping to
create a broad awareness and understanding of the
nature and importance of the AMS measurement capability. Secondly, because of controversy surrounding
the meaning of the radiocarbon result, measurement
aspects of artifact dating have been given intense
scrutiny. Such scrutiny is quite positive, for it gives the
possibility of added insight into unsuspected phenomena and sources of measurement uncertainty.
The Turin Shroud is believed by many to be the
burial cloth of Christ. The documented record, however, goes back only to the Middle Ages, to Lirey,
France (ca. 1353 AD) with the first firm date being
1357 AD when it was displayed in a Lirey church.

The Shroud of Turin

The radiocarbon dating of the Turin Shroud is

arguably the best known dating application of accelerator mass spectrometry, at least to the lay public. It
could not, or at least it would not have taken place without AMS, because most decay (beta) counting techniques would have consumed a significant fraction of
this artifact. Although still a destructive analytical technique, AMS required only "a postage stamp" amount of
the linen cloth (Ref. [20], Chap. 15). This particular
exercise is having a metrological impact well beyond
the radiocarbon date, per se. This is shown, in part, by
200

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

(b)

Fig. 15. The Turin Shroud. Shown in the montage are: (15a, upper left), the cover of the issue of Nature (16 February 1989) reporting the results
of the C measurements by AMS laboratories in Tucson, Zurich, and Oxford; and three singular features of the artifact: (15b, lower left),
the =50 mg dating sample received by the Tucson laboratory, showing the distinctive weave (3:1 herringbone twill), with dimensions about
1 cm X 0.5 cm; (15c, upper right), the characteristic negative image, considered by some as a remarkable piece of mediaeval art; and (15d, lower
right), a microphotograph by Max Frei showing individual fibers supporting pollen grains of presumed unique origin [38, 39].

201

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

Radiocarbon dating was seen immediately as a definitive method to decide whether the "Lirey Shroud"
could have come fi-om flax grown in the 1st century
AD. The Shroud image, considered by some to be the
skilled work of a mediaeval artist, shows a fiill length
image of a crucified man; but as a negative image [Fig.
15c].' Prior to the AMS measurements, the Shroud was
subject to intensive examination by photography, spectroscopy, art and textile analysis, and palynology [3840]. The unique herringbone twill [Fig. 15b] is considered consistent with a 1st Century date; and pollen
grains found on the cloth [Fig. 15d] are stated by Max
Frei to have originated Irom a plant found only in the
region of Jerusalem. Radiocarbon dating of the cloth,
however, yielded a result of 1262 to 1384 AD (95 %
confidence interval) [38].
Apart from sampling,'" the AMS measurements were
performed taking the strictest quality control measures.

Three highly competent laboratories were selected: the

University of Arizona, Oxford University, and the
Swiss Federal Institute of Technology [ETH] in
Zurich. Samples of the Shroud, plus three control
samples of known age, were distributed blind to the
three laboratories. Control of this operation (distribution of samples, collection of results) was the responsibility of Michael Tite of the British Museum. The
accuracy and precision of the interlaboratory data for
the control samples were outstanding, leaving no doubt
as to the quality of the AMS measurement technique
(Fig. 16). Sample-1 (Shroud) results, however, were
just marginally consistent among the three laboratories,
prompting the authors of Ref [38] to state that "it is
unlikely that the errors quoted by the laboratories for
sample-1 fully reflect the overall scatter." Consistent
with the discussion in Sec. 2, the "'C age measurements
are reported in "''*C years BR" Transformation of these
ages to calendar ages must take into account the natural "'C variations, using the dendrochronological
calibration curve [13]. The transformation is shown in
Fig. 17, which demonstrates also an interesting aspect
of the non-monotonic calibration function: namely,
exclusion of the period between 1312 AD and 1353 AD
from the 95 % confidence interval. In addition, an
interesting link exists between this figure and
Fig. 6 (Maunder Minimum), in that the same solaractivity-induced ''*C variations are represented. A comparison of the two figures shows that the radiocarbon date (691 BP), near the end of a significant
calibration curve protrusion (Fig. 17), corresponds to
the end of the 13 th century warm period having high
solar activity (Fig. 6).
Consistency of the AMS results with the existing
(Lirey) documentation seems compelling, but a wave
of questioning has followednot of the AMS method,
but of possible artifacts that could have affected the
linen and invalidated the ''*C result (Ref [40], Chap. 1,
Refs. [41], [42]). A sampling of the creative hypotheses
put forward is given in Table 2. The first, for example,
is based on the premise that nuclear reactions involving
the substantial amount of deuterium contained in a
human body could produce neutrons, which might then
produce excess ''*C through the (n,p) reaction, making
the age too young. The proposed deuteron reactions,
however, are either qualitatively or quantitatively
inaccuratebarring an unnatural burst of high energy
photons (photofission). The third proposal raises the
question of non-contemporaneous organic matter
whether from incompletely removed carbon contamination from "oil, wax, tears, and smoke" that the cloth
had been exposed to, or from bacterial attack and

Figure 15c and 15d images are from the documentary prepared by
the British Broadcasting Corporation which is hereby acknowledged
[39]; Fig. 15b is courtesy of D. J. Donahue.
The critical, non-AMS issue relates to sample validity. The originally agreed upon sampling protocol was to have involved seven
laboratories, two measurement techniques (decay and atom [AMS]
counting), and multiple samples representing different regions of the
cloth. Shortly before the event, however, the scheme was changed to
restrict the number of laboratories (all AMS) and the number of
samples to three, all taken from the same location. The sampling
location, near a comer of the Shroud, and near an area damaged by
the fire of 1532 AD, is considered an unfortunate choice, because
of the possibility of exogenous carbon from the fire, repairs, and
organic contamination from handling through the ages [40, 41].
Organic contamination cannot be dismissed. Recent observations
indicate the presence of a bacterially-induced "bioplastic" coating on
Shroud fibers, as has sometimes been found on mummy wrapping
fabric (leading to ertoneous dates). According to [42] (Gove, et al.),
such bioplastic contamination would not have been removed by the
conventional pre-treatment methods applied to the Shroud samples.
Qualitatively, such contamination would lead to a more recent date;
quantitatively, if the contamination were all from the 16th Century, it
would need to represent roughly 70 % of the carbon present, to shift
a first century date to the observed result. (For recent, late 20th
Century contamination, roughly 40 % contamination carbon would
be required.) In a 2002 review article posted to the shroud website,
www.shroud.com/pdfs/rogers2.pdf, [38], Rogers and Amoldi
question the bioplastic hypothesis, on the basis of detailed chemical
analysis of fibers from the "Raes sample" which was taken from a
region adjacent to that of the C samples. Quantitatively, these
authors suggest that the coating would contribute only a few percent
to the sample carbon; qualitatively, they believe that it is a polysaccharide gum (probably Gum Arabic) that would be removed by
the C pretieatment chemistry. Nevertheless, Rogers and Amoldi
question the validity of the C sample, partly because of the
presence of cotton and other chemical differences between the
adjacent (Raes) sample and the main shroud material.

202

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

D'Anjou cope
Cleopatra mummy
*

- Nubian tomb
Turin siiroud

4BB

3BB

1208

IBBB

2BBB

240B

Radiocarbon Age (yr BP)

Fig. 16. AMS C dating results ("blind") for the Turin Shroud (sample-1) and three control samples of known
age (samples-2,3,4), from the three AMS laboratories: Z (Zurich), O (Oxford), and A (Arizona). Dates are
expressed as "Radiocarbon Years" before present (BP); uncertainties represent 95 % confidence intervals [38].

Calendar age (BP)

1050

300 t
900

850

650

1100
1300
Calendar age IAD)

450

1500

Fig. 17. Transformation of the Radiocarbon Age (BP) to the Calendar

Age (AD) of the Shroud. The ^*C age (95% CI) of (691 31)
BP corresponds to a two-valued calendar age as a resuh of the nonmonotonic radiocarbon dating calibration curve. As indicated in the
figure, the projected calendar age ranges are: (1262-1312) AD and
(1353-1384) AD [38].

203

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

Table 2. Creative Hypotheses

soot and pollen, we have the possibility of controlling

the sample preparation blank to less than 0.2 |ag by
applying a "thermal discriminator" at a critical stage of
the process. Microgram level "^C soot studies have
already been successful in Greenland snow; and pollen
studies hold great promise for ice core dating, and perhaps even for dating the pollen foimd by Max Frei on
the Turin Shroud.'^ An important measurement issue
for ice core pollen relates to the amount needed for a
given dating precision. To give a rough estimate:
assuming 50 ng carbon per pollen grain, a pollen
age of 2000 years, and 5 % Poisson imprecision
(o=400 years); one would need to collect about 100
pollen grains. This might be accomplished in a few
hours, using the "hand picking" microscope technique
ofLongetal. [48].

Excess C from deuterium

spontaneous fission; cold tusion
C isotopic fractionation/ exchange (fire of 1532 AD)
biased sampling; "age" depends on location
Bioplastic coating; non-contemporaneous with linen
pretieatment chemistry

deposit over the ages. Apart from the effects of such

factors on the Shroud, the issue of organic reactions and
non-contemporaneous contamination of ancient materials can be a very serious and complex matter, deserving
quantitative investigation of the possible impacts on
measurement accuracy.'" Research questions of this
sort, including the classic problem of dating ancient
bone, form one of the key stimuli for the development
of "molecular dating"the topic of the following section.

7.

14

Table 3. Molecular Dating ( C AMS at the microgram level)

Dilution AMS quantifies 0.9 ng modem carbon (1999)

- soot/ pollen blank controllable to -0.2 ng (o = 60 ng)
- challenge: dating pure pollen grains from the Shroud

Fossil and biomass aerosol sources characterized in remote

atmosphere/ cryosphere (2.9 ng biomass soot quantified)

Individual amino acids dated in mammoth bones (LC/AMS)

Individual poly cyclic aromatic hydrocarbons dated in atmospheric particles and marine sediment (GC/AMS)

Emergence of ^-Molar "^C Metrology

Radiocarbon metrology is at the very moment in the

midst of still another revolution, involving the dating
(or isotopic speciation) of pure chemical fractions:
"molecular dating." For trace species, such as polycyclic aromatic hydrocarbons (PAHs), or remote, low
concentration samples, such as the soot or pollen in the
free troposphere or in ice cores, the sensitivity of AMS
is challenged to its ultimate. In order to understand the
nature of the challenge it is interesting to consider the
limiting factors. In a recent study it was shown that
10 % Poisson "error" (standard uncertainty) can be
achieved with 0.9 ng modem carbon, whereas machine
background is equivalent to 0.2 ng or less [43]. Sample
processing blanks, however, may range from 1 |ag to
15 |ag or more, and they may consist of both biomass
carbon and fossil carbon [44]. Thus, the ultimate limiting factor for very small sample AMS is the overall
isotopic-chemical blank. Environmental studies of '*
in individual chemical compounds can be successful at
the 1 Hg to 10 ng level, but only with stringent control
of the variability of the blank. This is in sharp contrast
with small sample, low-level counting where the
Poisson modem carbon limit (ca. 3 mg) and background limit (ca. 5 mg equivalent) far exceed the typical sample preparation blank (ca. 40 ng) [29]."
Some illustrations of pure compound "dating" by
NIST and collaborators are given in Table 3. The first
item refers to the aforementioned 1 |ag capability, using
"dilution AMS." For thermally stable species such as

There is a profound difference between background-limited decay

counting and blank-limited AMS, that may not be widely appreciated. Although the ultimate limitation in each case is "B" variability,
when B represents the instrumental background it tends to be reasonably well controlled, and under the best of circumstances, Poissonian
[45]. An extia degree of caution is needed, however, when the limiting "B" is an isotopic-chemical blank. At best, it might be assumed
normal; then replicate-based detection tests and confidence intervals
can be constructed using Student's-t. If the blank does not represent
a homogeneous or stationary state (as a reagent blank, well-mixed
environmental or biological compartment, etc.), such tests and intervals can be totally misleading. Non-stationary blanks may exhibit
(geochemically meaningful) structure, or they may be ertatic, reflecting a transient source of contamination [46].
12
Molecular dating" of the pure cellulose fraction of the Shroud, or
of the associated pollen, could furnish an interesting consistency test
for the published radiocarbon date. It would be especially interesting to put a "time stamp" on pollen whose point of origin has already
been ascribed to a location 10 km to 20 km east and west of
Jerusalem [47]. Such measurements are made feasible by the reduction of requisite sample sizes by a factor of ten or more, from what
AMS C dating required sixteen years ago. The question of noncontemporaneous fiber from 16th Century repairs, for example,
could be addressed by new C measurements on just 100 |xg of
fibers (=50, 1 cm linen fibers) from the main part of the Shroud. The
expected standard uncertainty would be equivalent to approximately
120 radiocarbon years ([43], Eq. 1).

204

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

7.1

Long-Range Transport of Fossil and Biomass

Aerosol

Ongoing multidisciplinary, multi-institutional research

on soot particles in remote and paleo-atmospheres,
which is absolutely dependent on the small sample
dating capability, is indicated in Fig. 18. The upper
portion of the figure relates to climate oriented research
on the sources and transport of fossil and biomass
aerosol to the remote Arctic [49]; the lower portion
relates to atmospheric and paleoatmospheric research at
Alpine high altitude stations and ice cores [50,51]. In the
remainder of this section we present some of the highlights and measurement challenges of the first project,
on the long-range transport of carbonaceous particles to
Summit, Greenland.
Cooperative research on this project, between NIST
and the Climate Change Research Center at the
University of New Hampshire (UNH), began in 1994.
It was catalyzed by the discovery of an unusually heavy
loading of soot on one of the air filters used for 'Be
sampling at Summit, Greenland by Jack Dibb of UNH
[52]. The Summit soot had been ascribed to the combination of intense boreal wildfire activity in the lower
Hudson's Bay region of Canada and exceptional atmospheric transport to central Greenland. Measurement of
''^C in the filter sample yielded definitive evidence for
biomass burning as the source of the soot. On one day
only (5 August 1994), the biomass carbon increased by
nearly an order of magnitude, with scarcely any change
in the fossil carbon concentration on the filter.
Supporting data for the origin of the biomass burning
carbon came from backtrajectory analysis, AVHRR
(infrared) satellite imagery of the source region, and
TOMS (ultraviolet) satellite imagery that was able to
chart the course of the soot particles from the source
wildfires to Summit. The several parts of this remarkable event are assembled in Figs. 19, 20 [29,49,52].'^
Since snow and ice can serve as natural archives for
atmospheric events, one may expect to find chemical

evidence of prior years' fire seasons in snowpits, fim,

and ultimately ice cores. This is illustrated in the upper
right portion of Fig. 18, which shows depth profile
sampling in a snowpit at Summit, overlaying an energy
dispersive spectrum and SEM image of a char particle
found near the 1994 fire horizon in a 1996 snowpit
[29]. An organic tracer of conifer combustion, methyl
dehydroabietate, was found also at the same depth [53].
Atmospheric science entered a new phase at Summit
during the "Winter-Over" project (1997-1998) [54].
For the first time, direct sampling of air and surface
snow took place over the polar winter, extending from
June 1997 to April 1998. A special achievement of
micro-molar "*C "dating" was the first seasonal data for
carbonaceous particles, deposited with the surface
snow.''* The seasonal record for biomass carbon
particles, shown in Fig. 21, was striking [55]. The large
spring peaks, in particular, consisted primarily of
biomass carbon: 0.76 (M = 0.03) modem carbon mass
fi-action (/M) for sample-1 (WOl), and 0.94 (0.01) mass
fraction for sample-8 (W08). Beyond the fossilbiomass apportionment, however, lay questions about
the nature and origin of the carbonaceous aerosol.
Especially intriguing are contrasts between the samples
showing summer [sample-4 (W04)] and spring
[sample-8 (W08)] biomass-C maxima in Fig. 21. To
explore these, a "multi-spectroscopic" approach was
taken, through which insights and supporting evidence
were derived from a variety of analytical techniques.
Results for one of the microanalytical techniques
employed, laser microprobe mass spectrometry
(LAMMS), are shown in Fig. 22.'^ The figure uses a
principal component projection to summarize multivariate (multi-mass) contrasts between the summer and
spring biomass peaks. It shows that the three summer
(W04) sub-samples tend to favor C" cluster ions
(n-even), typical of condensed carbon structure (and
graphite), whereas the three spring (W08) sub-samples
exhibit a more complex, oxygenated structure such as
occurs with biopolymers.

The observed (Fig. 20) vs inferred (Fig. 19) paths of the smoke
plume present an interesting contrast. The TOMS satellite image
shows the smoke approaching the southern tip of Greenland on 3, 4
August 1994 and departing toward Iceland on 6 August. The backtrajectory model employed in Ref [52] places the approach at a
somewhat higher latitude, and of course provides no departure
information.

The micro-molar C capability was essential for this work

because of the extremely small concentrations of particulate carbon
in the surface snow, especially during the winter (<10 ng C/ kg snow).
Microanalytical methods, such as LAMMS, are crucial for gaining
chemical insight on individual particles, or when only very small
snow (or ice) samples of remote aerosols are available, or needed for
high resolution studies. In contrast to the ng capability of the most
sensitive bulk analysis techniques, LAMMS, can provide useful
chemical data on as little as 20 pg of carbon species [57].

205

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

Summit, Greenland
- Particle (90% C) from
1994 fire horizon
- Biomass-C aerosol seasonal
cycle reported (2003)
[6 to 40 |ig/kg snow]
Mt. Sonnblick
Austria (3106 m)
Global Atmospheric
Watch Observatory
-''*C "dating" of soot in the free
troposphere [Weissenbok, 2000]

Fig. 18. Submicromolar C apportionment of anthropogenic and natural carbonaceous aerosols at remote sites in Europe and Greenland provides
knowledge of their impacts on present and paleoclimate [49-51].

206

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

207

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

^
fe

"^

^ o

> Z
'3 s

O w
^ o

B "
^ o
Si'o

^I
'o .2?

s
O T3

'15 .a

> ^
o 6
8

Q 3

208

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

Summit Snow -- Winter Over (1997-1998)

50

40 -

Spring ^
spring

/
/
/
/
/
/
/

c 30
o

o
August

E
o
ffl

/
/
10
Summer
Winter

4
5
Sample
Fig. 21. First evidence of a seasonal pattern in biomass carbon aerosol in surface snow in central Greenland
[55, 56]. Fundamental differences were found between the biomass carbon peaks in summer (sample-4 [W04]),
and spring (sample-8 [W08]) via "multi-spectroscopic" macro- and micro-chemical analysis.

Findings from other techniques:

Thermal-optical analysis. Distinctive seasonal

volatilization/decomposition patterns were seen as
samples were heated in a stream on helium. The
summer sample (W04) had a predominant high
temperature peak at =560 C and little evidence of
charring (4 %), whereas the spring sample (W08)
had a predominant peak at =410 C and major
charring (19 %). Thermal analysis of a powdered
wood (oak) reference material showed a thermal
peak at the approximately same temperature as
W08, with 21 % charring, implying the presence
of a major cellulosic component in this sample.
Ion chromatography. Fire tracers (NH/, K^)
accompanied W04; soil tracers (Ca^, Mg^)
accompanied W08
Backtrajectories. For W04, strong transport was
indicated from regions of annual wildfires in the
Canadian Northwest; for W08, strong transport
was indicated from the agricultural regions of the
upper Midwestboth representing transport
distances of some 8 Mm.

Electron probe microanalysis. For W04, up to

90 % C (mass fraction) was observed in individual, \im size particles, with C > O for the most
abundant (core) particles; for W08, maximum C
particles had a C:0 ratio consistent with cellulosic
biopolymer, and C < O for the core particles.

The weight of multi-spectroscopic evidence thus

indicates that the summer (W04) and spring (W08)
biomass particles do not represent the same type of
biomass. Rather, the W04 particles appear to include a
soot component from high temperature combustion
(motor vehicles, wildfires). The W08 particles, whose
carbon derives almost entirely from biomass, appear to
have a major biopolymer component, such as cellulose
and other bio-materials associated with soil and vegetative carbon. These findings are consistent with work by
Puxbaum and colleagues, who have found by direct
chemical analysis, significant amounts of cellulose,
bacteria, and fiingal spores in atmospheric particles [58,
59].
209

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

3 -

2 -

1 -

o
c
o

8 -

C2H2O-

-2 -

-3 _t

_i

-2

-1

W04

a
Conponent 1

i_

12

W08

Fig. 22. PCA biplot of laser microprobe mass spectral data; compositional contrast between particles
from the summer biomass peak (W04, red: C~ cluster ions favored) and the spring biomass peak
(W08, green: oxygenates favored) [55].

7.2

Isotopic Speciation in Ancient Bones and

Contemporary Particles

and the collagen hydrolysates [XAD-HYD]), age

concordance among the individual amino acids
and with the archaeological evidence indicated reliability. Had contamination from bio-intrusive material
having a different chemical (amino acid) pattern
occurred, amino acid age heterogeneity would have
been expected [60]. This work could not have been
accomplished without the ability to date 80 |ag carbon
fractions.
An historical footnote related to this work involves
the question of the ancestors of the North American
Clovis culture. Since the Clovis sites give the earliest
unequivocal data on the "peopling" of the Americas,
it has been of enormous interest to find a geochronological link to an earlier culture. The most popular
belief that the Clovis progenitors had arrived over
the "Bering Land Bridge" from Siberia has recently
been put into doubt, however, with new ''*C evidence
that one of the most likely pre-Clovis sites in northeastem Siberia is 4000 years younger than previously
believed. Dating at =13 000 calendar years ago, it is
doubtful that migration could have transpired quickly
enough to give rise to the Clovis culture (13 600 to
12 600 calendar years BP) in the North American
Southwest [61].

The dating of ancient bones has been notably

unreliable because of diagenesis and isotopic contamination that occur with millennia of environmental
exposure. Molecular dating of individual amino acids
in such bones has proven to be one of the most effective means to overcome this problem. Figure 23 shows
dramatically how the apparent radiocarbon age of
the Dent Mammoth changed from ca. 8000 BP to
ca. 11 000 BP, as the dated chemical fraction was
refined from the crude collagen fraction to the individual amino acids. The known radiocarbon age is given as
=11,000 BP, based on association with Clovis culture
artifacts, and biostratigraphy [60]. (Note that the
"calibrated" or corrected calendar age, derived from
the radiocarbon calibration curve [13], is roughly
1500 years older than the radiocarbon age for this time
period.) The commonly dated organic fractions from
bones (weak acid insoluble collagen [COLL] and
gelatin [GEL]) gave ages that were at odds with the
archaeological evidencesuggesting recent humate
contamination. When the diagenesis-resistant molecular components were isolated (individual amino acids
210

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

8 -t

9 -

O
X
LU

>

a
<

O
O

10

LU
CD

<

(aliphatic extract) to 38 % (total carbon). Thus, the

aliphatic fraction derives essentially (=98 %) from
fossil fuel emissions, and, on average, fossil sources
account for some 60 % of the carbon in these particles.
Note that the Certificate of Analysis [63] provides
"*C data expressed in the proper reference units as
fraction of modem carbon (Z^). To emphasize the more
meaningful fossil-biomass carbon source dichotomy,
however, we have chosen to present the information
here in terms of the fraction of biomass carbon.
Conversion is based on the "post-bomb" enrichment of
''^C in the living biosphere, as shown in Fig. 11.
Sampling for SRM 1649a took place in 1976-1977;
the enrichment factor for biomass carbon at that time,
indicated by the red arrow in the figure, was 1.35.
One of the most important outcomes of the SRM
1649a intercomparison exercise was the set of data
obtained for "elemental carbon" (EC). EC (sometimes
known as "black carbon") is routinely monitored in
urban and rural aerosols, and it is of major concern
because of its presumed impacts on health, visibility,
and climate (radiation absorption). SRM 1649a
potentially can serve as a key laboratory quality
assurance reference material for EC measurement.
Results of the largest intercomparison to date of EC in
a uniform reference material, however, indicate a
severe measurement problem: relative values for the
reported data span a range of 7.5, showing very significant method dependence. Three clusters of results
for the mass fraction of EC (relative to total-C), reported
as information values on the Certificate of Analysis,
are 0.075, 0.28, and 0.46. (For the ''C data in Fig. 24,
cluster-1 EC has been labeled "soot" and cluster-3 EC,
"char." ''*C was not determined in cluster-2 EC.)
The fiindamental problem is that EC is not a pure
substance, so a unique "true value" for EC may not
exist, in principle."' Some interesting insights into the
meaning of certain of the EC results follow, however,
from the ''*C EC speciation data.

DENT
MAMMO I II

LU

X
I

Q.

ir

CL

>.

<

>-

-I

_i

1-

C3

<

<

C5

T" T

o
11 -

"~I

SAMPLE
Fig. 23. "Molecular Dating" of individual amino acids in ancient
bone. Radiocarbon ages of commonly dated (collagen, gelatin)
fractions were 2000 to 3000 years too young as a result of environmental degradation; pure molecular fractions (amino acids) were
self-consistent and in agreement with the Clovis culture age [60].

7.2.1

Urban Dust (SRM 1649a): a Unique

Isotopic-Molecular Reference Material

SRM 1649a is NIST's most highly characterized

natural matrix Standard Reference Material, and it is
the only one for which there are certificate values for
"*C in individual chemical fractions and pure molecular
species. The "carbon" portion of the Certificate of
Analysis was developed through an extensive international interlaboratory comparison, involving eighteen
teams of analytical experts from eleven institutions
[62]. The particle-based SRM, which has been characterized for nearly 200 chemical species and properties,
serves as an essential quality assurance material for a
remarkably broad range of disciplines, from the monitoring of pesticides, PCBs, and particulate mutagenic
activity to basic organic geochemistry to isotopic
apportionment of carbonaceous particles. A dramatic
illustration of the '* isotopic heterogeneity in this
reference material is given in Fig. 24. The biomass
carbon mass fractions are seen to range from about 2 %

Although a "true" (Certified) EC value may be beyond reach, compatibility of results from laboratories using the same method suggests
the possibility of method-specific ("operational") EC Reference
Values for this SRM.

211

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

Urban Particulate Reference Material (SRM 1649a)

(prototypical isotopic-chemical aerosol reference/QA material)
i^'C SPECIATION
CARBON

BIOMASS-C (%

total

38

polar

32

elemental
"char"

11

"soot"

aromatic

13

aliphatic

Pyrene

Benzo(g/;/)perylene

(U = 6 [aromatic]; others <1)

Fig. 24. NIST Standard Reference Material 1649a ("urban dust"). Photograph of the bulk reference material and derived "filter samples" for QA
of atmospheric elemental carbon (EC). C data listed indicate the mass fraction (%) of biomass-C in the several chemical fractions [29, 62].

Regarding the second test, the "*C data in Fig. 24

demonstrate that isotopic-mass balance cannot be
achieved with the current isotopic-chemical data. Since
the biomass carbon fraction on average (38 % mass
fraction) exceeds that of all other measured fractions,
there must be a significant missing biomass carbon
component. This matter is addressed in [62], where it is
suggested that unmeasured biopolymers may account
for more than 45 % of the residual (non-extractable,
non-EC) carbon mass. Cellulose is one excellent candidate [58].

Isotopic consistency. Measurement of '* in multiple

chemical fractions offers the possibility of two very
interesting and important consistency tests: (1) assessment of isotopic-chemical consistency among chemically-related fractions, and (2) assessment of overall
isotopic-mass balance. The first test is illustrated by
comparison of the '* content of the EC fraction with
that of the PAH fraction (on average). To the extent that
both components originate from the same source,
acetylenic free radicals that generate polyaromatic
structures in the flaming stage of combustion, one
would expect similar '* composition. Such is the case
for "*C in cluster-1 EC (labeled "soof in Fig. 24), but
not for cluster-3 EC (labeled "char"). The lack of
isotopic consistency for cluster-3 EC is the stimulus for
the different label, since this manifestation of EC
necessarily reflects a different mix of fossil-biomass
sources than the flaming stage EC, which derives
primarily from fossil fuel carbon.

GC/AMS. Finally, the "molecular dating" of individual

PAH in SRM 1649a epitomizes one of the latest
advances in micromolar ''*C measurement science: the
capability to link chromatographic isolation of pure
chemical compounds to AMS determination of ''^C'^C.
Results of applying off-line GC/AMS to six PAHs
recovered from the aromatic fraction of SRM 1649a are

212

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

% Biomass-C
\..
'-

'

T-

r^ 1

-f

I-

~l

3.0

^t^^^^^^^^^

'i '^T

_L

I "

4.7

-"

2.8

\,

-I-^

-<-I-

j'

u
I

3.1

6.2

6.4

JWW*--^

-_L^>_^fc>< t t\, f >*vi|

30

40

50

60

70

80

90

Fig. 25. Gas chromatography/accelerator mass spectrometry (GC/AMS): AMS following automated
prep-scale capillary GO yields "dates" (equivalent biomass carbon mass fractions) for micromolar
amounts of individual polycyclic aromatic hydrocarbons [63-65]. (Results shown forNIST SRM 1649a;
"I.S." denotes an internal standard; abscissa indicates retention time (min).)

shown in Fig. 25. The critical first step was the sequential isolation of tens of micrograms of the six PAHs in
separate traps by automated preparative scale capillary
gas chromatography [66]. The individually trapped
PAHs were then oxidized and converted to AMS targets. These results represent the first such data ever
available for an atmospheric particulate SRM, and
although such compounds are only trace constituents of
atmospheric particles (=10 ng/g), they are of great consequence due to their mutagenic and carcinogenic properties. In this case, as shown in Fig. 25, radiocarbon
dating of the individual PAHs revealed these congeners
to be isotopically heterogeneous, and demonstrated a

basic flaw in the conventional wisdom that the heavier

PAHs, in particular, are more likely to be produced
strictly from fossil fuel combustion sources.
On-line GC/AMS is nearly upon us. The linkage of
gas (or liquid) chromatographic separation, and direct
injection of microgram amounts of pure compounds
into the ion source of an accelerator mass spectrometer,
is under active investigation in several AMS laboratories; and it promises a new dimension in the practice of
radiocarbon dating at the molecular level that may have
an impact on archaeology and isotopic biogeochemistry comparable to that of GC/MS on analytical,
physical, organic, and biochemistry [67].
213

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

8.

Epilogue

courtesy of Douglas J. Donahue, University of Arizona.

Fig. 6a (top), reprinted with permission from Fig. 5a in:
Eddy, J.: "The Maunder Minimum," Science 192
(1976) 1189-1202; copyright 1976 American
Association for the Advancement of Science. Fig. 6b
(bottom), from the "climate" figure (p. 615, last segment only, labeled "Past 1000 years") in: Mathews, S.:
"What's happening to our climate," National
Geographic 150 (1976) 176-615; copyright 1976 the
National Geographic Society. Fig. 7 and Fig. 8, from:
Toggweiler, J.R., Dixon, K. and Bryan, K.:
"Simulations of Radiocarbon in a Coarse-Resolution
World Ocean Model, 2. Distributions of bomb-produced "'C," J Geophys Res 94 [C6] (1989) 8243-8264
(figures 1 and 17, respectively); copyright 1989
American Geophysical Union. Figures 9 and 10 are
adapted from Currie, L.A., et al., "Authentication and
Dating of Biomass Components of Industrial Materials:
Links to Sustainable Technology," Nuclear Instruments
and Methods in Physics Research B172 pp 281-287,
copyright (2000), with permission from Elsevier
Science. Fig. 12: photos are courtesy of Robert K.
Stevens. Fig. 13, from Fig. 1 in: Wolfli, W: "Advances
in accelerator mass spectrometry," Nuclear Instruments
and Methods in Physics Research B29 [numbers 1, 2]
pp 1-13, copyright (1987), with permission from
Elsevier Science. Fig. 14, from Fig. 16.2 in: Taylor,
R.E., Long, A., and Kra, R., Eds.: Radiocarbon after
Four Decades: an Interdisciplinary Perspective; copyright Springer-Verlag, New York, 1992. Fig. 15c inset
(negative image) and 15d (microphotograph), from:
British Broadcasting Corporation Documentary,
"Shreds of Evidence" (Timewatch Series), copyright
1988. Fig. 15b, courtesy of Douglas J. Donahue,
University of Arizona. Figures 15a (reprint cover), 16,
and 17, from Damon, P., et al.: "Radiocarbon dating of
the Shroud of Turin," Nature 337 (1989) 611-615;
copyright Macmillan Magazines Ltd, 1989. Fig. 18,
from: (a) Currie, L.A., et al.: "The pursuit of isotopic
and molecular fire tracers in the polar atmosphere and
cryosphere," Radiocarbon 40 (1998) 381-390, 416f;
copyright 1998 Arizona Board of Regents on behalf of
the University of Arizona; and (b) Mark Twickler,
Univ. New Hampshire. Fig. 19 (center, backtrajectories), from Dibb, J.E., et al., "Biomass burning signatures in the atmosphere and snow at Summit,
Greenland: an event on 5 August 1994," Atmospheric
Environment 30 pp 553-561 copyright (1996),
with permission from Elsevier Science. Figures 19,
20 adapted from Currie, L.A., et al., "The pursuit of
isotopic and molecular fire tracers in the polar atmos-

Libby's discovery, and the remarkable developments

that followed, arose from a scientific question (freely
translated): "What will become of the cosmic ray
neutrons?" It is noteworthy that an "academic son" of
this Nobel Laureate also posed a scientific question to
himself. F. Sherwood Rowland's question also led to an
unexpected discovery having major practical import for
mankind: the possible destruction of the stratospheric
ozone layer. Rowland's query, also culminating in a
Nobel Prize (1995), was "I began to wonder what
was going to happen to this man-made compound
[trichlorofluoromethane] newly introduced into the
atmosphere" [68].
May this historical journey into scientific discovery,
as an outgrowth of seemingly simple scientific curiosity, and the consequent unanticipated scientific-metrological revolutions, encourage students to examine the
original historical literature documenting such discoveries, and to realize that profound unforeseen developments may be in store for a presumably "mature"
scientific discipline.
Acknowledgment
This article represents an adaptation and extension of
a recent publication in the Czechoslovak Journal of
Physics: "The Remarkable Metrological History of '*
Dating: from ancient Egyptian artifacts to particles of
soot and grains of pollen" [Czech. J. Phys. 53, (Suppl.
A) A137-A160 (2003)]. Permission of the Institute of
Physics, Academy of Sciences of the Czech Republic is
gratefully acknowledged. Thanks go also to Cynthia
Zeissler and Ed Mai for assistance in final preparation
of the figures for publication.
Figures are adapted, with permission, from the
following sources. Fig. 1: photo by Fabian Bachrach
(AEC-54-5123-DOE) from page 1 of de Messieres, N.:
"Libby and the interdisciplinary aspect of radiocarbon
dating." Radiocarbon 43 (2001) 1-5; copyright 2001
Arizona Board of Regents on behalf of the University
of Arizona. Fig. 3, from Radiocarbon Dating [jacket
cover] (Eds. R. Berger and H. Suess) Univ. California
Press, Berkeley, 1979]. Fig. 4, from Fig. 1 of: Libby,
Willard F., Radiocarbon Dating, Univ. Chicago Press,
Chicago, copyright 1952 (1st edition). Cover and Fig. 5
(plot), from Fig. 1 (p. 110) in: Olsson, I.U., Ed.
Radiocarbon Variations and Absolute Chronology (12th
Nobel Symposium), Almqvist & Wiksell, Stockholm,
1970; copyright, the Nobel Foundation. Fig. 5 (photo).

214

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

[12] I. U., Olsson, ed.. Radiocarbon Variations and Absolute
Chronology (12th Nobel Symposium), Almqvist & Wiksell,
Stockholm (1970).
[13] M. Stuiver and R. Kra, eds.. Calibration Issue, Radiocarbon 28,
2B (1986); Stuiver, M., Long, A., Kra, R., eds.. Calibration
1993, Radiocarbon 35 (1), 191-199 (1993).
[14] M. Geyh and C. Schluchter, Calibration of the ''^C time scale
beyond 22,000 BP, Radiocarbon 40, 475-482 (1998).
[15] J. Eddy, The Maunder Minimum, Science 192, 1189-1202
(1976). (See Fig. 3).
[16] S. Mathews, What's happening to our climate. National
Geographic 150, 176-615 (1976). (See especially pp. 586,
614 f).
[17] J. Lopes, R. Pinto, M. Almendra, and J. Machado, Variation of
C activity in Portuguese wines from 1940 to 1974, Proc. Int.
Conf on Low-Radioactivity Measurements and Applications:
The High Tatras, Czechoslovakia, 6-10 October 1975, P.
Povinec and S. Usacev, eds., Comenius Univ., Bratislava (1977)
pp. 265-268.
[18] W. F. Cain, C in Modem American Trees, in Radiocarbon
Dating, R. Berger and H. Suess, eds., Univ. California Press,
Berkeley (1979) pp. 495-510.
[19] J. R. Toggweiler, K. Dixon, and K. Bryan, Simulations of
Radiocarbon in a Coarse-Resolution World Ocean Model, 2.
Distributions of bomb-produced C, J. Geophys. Res. (C6) 94,
8243-8264 (1989). [See also: I. Levin and V. Hesshaimer,
Radiocarbona unique tracer of global carbon cycle
dynamics. Radiocarbon 42, 69-80 (2000).]
[20] R. E. Taylor, A. Long, and R. Kra, eds.. Radiocarbon After Four
Decades: An Interdisciplinary Perspective, Springer-Verlag,
New York (1992).
[21] M. Stuiver and G Ostlund, GEOSECS Atlantic Radiocarbon,
Radiocarbon 22, 1-24(1980).
[22] J. Sarmiento, and N. Gruber, Sinks for anthropogenic
carbon. Physics Today 55, 30-36 (8) (2002). [See also: R.
Nydal, Radiocarbon in the Ocean, Radiocarbon 42, 81-98
(2000).]
[23] A. P McNichol, R. J. Schneider, K. F von Reden, A. R.
Gagnon, K. L. Elder, NOSAMS, R. M. Key and P D. Quay, Ten
years after: The WOCE AMS radiocarbon program, Nucl.
Instrum. Meth. Phys. Res. B172, 479-484 (2000).
[24] E. M. Wild, K. A. Arlamovsky, R. Golser, W. Kutschera, A.
Priller, S. Puchegger, W. Rom, P Steier, and W. Vycudilik, ''*C
dating with the bomb peak: an application to forensic
medicine, Nucl. Instrum. Meth. Phys. Res. B172, 944-950
(2000).
[25] L. Currie, D. Klinedinst, R. Burch, N. Feltham, and R. Dorsch,
Authentication and Dating of Biomass Components of
Industrial Materials: Links to Sustainable Technology, Nucl.
Instrum. Meth. Phys. Res. B172, 281-287 (2000).
[26] G Clayton, J. Arnold, and F. Patty, Determination of Sources of
Particulate Atmospheric Carbon, Science 122, 751-753(1955).
[27] J. P Lodge, G S. Bien, and H. E. Suess, The Carbon-14 Content
of Urban Airborne Particulate Matter, Int. J. Air Pollut. 2, 309
(1960).
[28] T. H. Maugh, Air pollution: where do hydrocarbons come
trom? Science 189, 277 (1975).
[29] L. A. Currie, Evolution and Multidisciplinary Frontiers of C
Aerosol Science, Radiocarbon 42, 115-126 (2000).

phere and cryosphere," Radiocarbon 40 (1998) 381390,416f; copyright 1998 Arizona Board of Regents on
behalf of the University of Arizona. Fig. 23, from
Fig. 1 in: Stafford, T.W., Jr., Hare, RE., Currie, L.A.,
Jull, A.J.T., and Donahue, D.: "Accuracy of North
American Human Skeleton Ages", Quarternary
Research, 34, pp 111-120, copyright (1990), with
permission from Elsevier Science. Fig. 24, adapted
from Fig. 4 in Currie, L.A., "Evolution and
Multidisciplinary Frontiers of '* Aerosol Science,"
Radiocarbon 42 (2000) 115-126, copyright 2000
Arizona Board of Regents on behalf of the University
of Arizona. Fig. 25, from Fig. 2 in: Currie, L.A.,
Klouda, G.A., Benner, Jr., B.A., Garrity, K., and
Eglinton, T.I., "Isotopic and Molecular Fractionation in
Combustion; Three Routes to Molecular Marker
Validation, including Direct Molecular 'Dating'
(GC/AMS)," Atm. Environ. 33 (1999) 2789-2806; published by Elsevier Science Ltd.

9.

References

[1] W. F. Libby, Nuclear dating: an historical perspective, Nuclear

and Chemical Dating Techniques: Interpreting the
Environmental Record, L. A. Currie, ed., American Chemical
Society Symposium Series No. 176 (1982) Chap. 1.
[2] Willard F. Libby, Radiocarbon Dating, Univ. Chicago Press,
Chicago (1952).
[3] N. de Messieres, Libby and the interdisciplinary aspect of
radiocarbon dating. Radiocarbon 43, 1-5 (2001).
[4] W. F. Libby, Radiocarbon dating (Nobel Lecture), Science 133,
621-629 (1961).
[5] W. F. Libby, Atmospheric Helium-3 and Radiocarbon from
cosmic radiation, Phys. Rev. 69, 671-672 (1946).
[5a] H. Godwin, Half-life of radiocarbon. Nature 195, 984 (1962).
[6] E. C. Anderson, W. F. Libby, S. Weinhouse, A. F. Reid,
A. D. Kirshenbaum, and A. V. Grosse, Radiocarbon from
cosmic radiation. Science 105, 576 (1947). (See also C.
MacKay, M. Pandow, and R. Wolfgang, On the chemistry of
natural radiocarbon, J. Geophys. Res. 68, 3929-3931 (1963),
for an account of CO as the precursor of cosmogenic CO2.)
[7] E. C. Anderson and W. F. Libby, World-wide distribution of
natural radiocarbon, Phys. Rev. 81, 64-69 (1951).
[8] J. R. Arnold and W. F. Libby, Age determination by radiocarbon content: checks with samples of known age. Science
110,678-680(1949).
[9] I. Karlen, I. Olsson, P. Kallberg, and S. Killici, Absolute determination of the activity of two C dating standards. Ark.
Geotys. 4, 465 (1964). (See also Ref [12], p. 17, The use of
oxalic acid as a [radiocarbon dating] standard.)
[10] NIST Standard Reference Materials website: http://nist.gov/srm.
[11] K. Rozanski, W. Stichler, R. Gonfiantini, E.M. Scott, R.
Beukens, B. Kromer, and J. van der Plicht, The IAEA intercomparison exercise 1990, Radiocarbon 34, 506-519 (1992).

215

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

[30] R. K. Stevens, C. W. Lewis, T. G. Dzubay, R. E. Baumgardner,
R. B. Zweidinger, V. R. Highsmith, L. T. Cupitt, J. Lewtas,
L. D. Claxton, L. A. Currie, G. A. Klouda, and B. Zak,
Mutagenic Atmospheric Aerosol Sources Apportioned by
Receptor Modeling, ASTM Monograph STP 1052, Monitoring
Methods for Toxics in the Atmosphere, W. L. Zielinski, Jr. and
W. D. Dorko, eds., Amer. Soc. Testing Mtls., Philadephia
(1990) pp. 187-196.
[31] D. B. Klinedinst, and L. A. Currie, Direct Quantification of
PM2.5 Fossil and Biomass Carbon within the Northern Front
Range Air Quality Study's Domain, Environ. Sci. Technol. 33,
4146-4154(1999).
[32] G A. Klouda, C. W. Lewis, D. C. Stiles, J. L. Marolf,
W. D. Ellenson, and W. A. Lonneman, Biogenic contributions
to atmospheric volatile organic compounds in Azusa,
California, J. Geophys. Res. 107 (D8) 10.1029/2001JD000758
(2002).
[33] T. Kuc and M. Zimnoch, Changes of the CO2 sources and sinks
in a polluted urban area (Southern Poland) over the last decade,
derived from the carbon isotope composition. Radiocarbon 40,
417-423 (1998).
[34] R. A. Muller, Radioisotope dating with a cyclotron. Science
196,489-494(1977).
[35] D. Nelson, R. Korteling, and W. Stott, Carbon-14: Direct
detection at natural concentrations. Science 198, 507-508
(1977).
[36] C. L. Bennett, R. P Beukens, M. R. Clover, H. E. Gove, R. B.
Liebert, A. E. Litherland, K. H. Purser, and W. E. Sondheim,
Radiocarbon dating using electrostatic accelerators: negative
ions provide the key. Science 198, 508-510 (1977).
[37] W. Wolfli, Advances in accelerator mass spectrometry, Nucl.
Instrum. Meth. Phys. Res. B29, 1-13 (1987).
[38] P Damon, D. J. Donahue, B. H. Gore, A. L. Hatheway, A. J. T.
Jull, T. W. Linick, P J. Sercel, L. J. Toolin, C. R. Bronk, E. T.
Hall, R. E. M. Hedges, R. Housley, I. A. Law, C. Perry, G
Bonani, S. Trumbore, W. Wolfli, J. C. Ambers, S. G E.
Bowman, M. N. Leese, and M. S. Tite, Radiocarbon dating of
the Shroud of Turin, Nature 337, 611-615 (1989). (See also the
website: www.shroud.com/papers.htm for this article
(.../nature.htm) and others contained in a large on-line collection of scientific articles treating various aspects of Shroud
research.)
[39] British Broadcasting Corporation: Documentary, Shreds of
Evidence, Timewatch series (1988).
[40] H. Gove, Relic, Icon, or Hoax? Carbon Dating the Turin
Shroud, Institute of Physics Publishing, Bristol and
Philadelphia (1996).
[41] I. Wilson, An appraisal of the mistakes made regarding the
Shroud samples taken in 1988, Proc. 3rd Intern. Conf on the
Shroud of Turin. Turin, June 1998. (www.shroud.com/wilson.htm).
[42] Articles examining non-contemporaneous organic contamination of textiles: 1) H. Gove, S. Mattingly, A. David, and
L. Garza-Valdes, A problematic source of organic contamination of linen, Nucl. Instixim. Meth. Phys. Res. B123, 504-507
(1997) ; 2) A. Jull, D. Donahue, and P Damon, Factors that
affect the apparent radiocarbon age of textiles. Archaeological
Chemistry, M. Orna, ed., ACS symposium series 625,
American Chemical Society, Washington DC (1996) Chap. 19,
pp. 248-253.
[43] L. A. Currie, J. D. Kessler, J. V. Marolf, A. P McNichol, D. R.
Stuart, J. C. Donoghue, D. J. Donahue, G .S. Burr, and D.
Biddulph, Low-level (submicromole) environmental
C

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

216

meh-ology, Nucl. Instrum. Meth. Phys. Res. B172, 440-448

(2000).
D. B. Klinedinst, and L. A. Currie, Radiocarbon blank correction: Methodologies and limitations in a major urban study of
carbonaceous aerosols, Nucl. Insttum. Meth. Phys. Res. B172,
545-550 (2000).
L. A. Currie, E. M. Eijgenhuijsen, and G A. Klouda, On the
validity of the Poisson Hypothesis for low-level counting;
investigation of the distributional characteristics of background
radiation with the NIST Individual Pulse Counting System,
Radiocarbon 40, 113-127 (1998).
L. A. Currie, Some case studies of skewed (and other abnormal) data distoibutions arising in low-level environmental
research, Fresenius J. Anal. Chem. 370, 705-718 (2001).
A. Danin, and U. Baruch, Floristic indicators for the origin of
the Shroud of Turin, Proc. 3rd Intern. Conf on the Shroud of
Turin. Turin, June 1998. (www.shroud.com/pdfs/daninx.pdf).
A. Long, O. Davis, and J. de Lanois, Separation and C dating
of pure pollen from lake sediments: nanofossil AMS dating,"
Radiocarbon 34, 557-560 (1992).
L. A. Currie, J. E. Dibb, G A. Klouda, B. A. Benner, Jr., J. M.
Conny, S. R. Biegalski, D. B. Klinedinst, D. R. Cahoon, and
N. C. Hsu, The pursuit of isotopic and molecular fire tracers in
the polar atmosphere and cryosphere. Radiocarbon 40, 381390, 416f (1998). [See also: N.C. Hsu, J. R. Herman, P .K.
Bhartia, C. J. Seftor, 0. Torres, A. M. Thompson, J. F. Gleason,
T .F. Eck, and B. N. Holben, Detection of biomass burning
smoke fiom TOMS measurements, Geophys. Res. Lett. 23,
745-748 (1996).
R. Weissenbok, L. A. Currie, C. Grollert, W. Kutschera, A.
Priller, H. Puxbaum, W. Rom, P. Steier, and J. Marolf,
Accelerator Mass Spectrometry Analysis of Non-soluble
Carbon in Aerosol Particles from High Alpine Snow (Mt.
Sonnblick, Austoia), Radiocarbon 42, 285-294 (2000).
CARBOSOL: Present and retrospective state of organic vs
inorganic aerosol over Europe: implications for climate,
European Union; M. Legrand, coordinator, (2001-).
J. E. Dibb, R. W. Talbot, S. I. Whitiow, M. C. Shipham, J.
Winterle, J. McConnell, and R. Bales, Biomass burning signatures in the atmosphere and snow at Summit, Greenland: an
event on 5 August 1994, Atmospheric Environment 30, 553561 (1996).
J. F. Slater, L. A. Currie, J. E. Dibb, and B. A. Benner, Jr.,
Distinguishing the relative contribution of fossil-fuel
and biomass combustion aerosols deposited at Summit,
Greenland, through isotopic and molecular characterization of
insoluble carbon. Atmospheric Environment 36, 4463-4477
(2002).
J. E. Dibb, convenor, American Geophysical Union Special
Session: Relationships Between the Composition of the
Atmosphere and Snow, AGU Fall Meeting (San Francisco,
December 1998).
L. A. Currie, J. D. Kessler, R. A. Fletcher, and J. E. Dibb,
Long-range transport of biomass aerosol to Greenland: multispectroscopic investigation of particles deposited in the snow, J
Radioanal. Nucl. Chem. 263 (2005) in press.
J. D. Kessler, L. A. Currie, E. S. Windsor, and D. E. Newbury,
Seasonal record of total particulate matter embedded in
Greenland surface snow, Microsc. Microanal. 7, 478-479
(2001).
R. Fletcher, R. Dobbins, and H.-C. Chang, Mass Spectrometry
of particles Formed in a Deuterated Ethene Diffusion Flame,
Anal. Chem. 70, No. 13, 2745-2749 (1998).

Volume 109, Number 2, March-April 2004

Journal of Research of the National Institute of Standards and Technology

[58] H. Puxbaum, and M. Tenze-Kunit, Size distribution and
seasonal variation of atmospheric cellulose, Atmospheric
Environ. 37, 3693-3699 (2003).
[59] H. Bauer, A. Kasper-Giebl, M. Loflund, H. Giebl, R.
Hintzenberger, F. Zibuschka, and H. Puxbaum, The contribution of bacteria and fungal spores to the organic carbon content
of cloud water, precipitation and aerosols. Atmospheric
Research 64, 109-119 (2002).
[60] T. W. Stafford, Jr., P. E. Hare, L. A. Currie, A. J. T. Jull, and D.
Donahue, Accuracy of North American Human Skeleton Ages,
Quartemary Res. 34, 111-120 (1990).
[61] T. Goebel, M. R. Waters, and M. Dikova, The archaeology of
Ushki Lake, Kamchatka, and the Pleistocene peopling of the
Americas, Science 301, 501-505 (2003).
[62] L. A. Currie, B. A. Benner, Jr., H. Cachier, R. Gary, J. C. Chow,
E. R. M. Druffel, T. I. Eglinton, O. Gustafsson, P C. Hartmann,
J. I. Hedges, J. D. Kessler, T. W. Kirchstetter, D. B. Klinedinst,
G A. Klouda, J. V. Marolf, C. A. Masiello, T. Novakov, A.
Pearson, K. M. Prentice, H. Puxbaum, J. G Quinn, C. M.
Reddy, H. Schmid, J. F. Slater, J. Watson, and S. A. Wise, A
Critical Evaluation of Interlaboratory Data on Total, Elemental,
and Isotopic Carbon in the Carbonaceous Particle Reference
Material, NIST SRM 1649a, J. Res. Natl. Inst. Stand. Technol.
107, 279-298 (2002).
[63] NIST SRM 1649a (Urban Dust) Certificate of Analysis, Jan.
2001 (http://nist.gov/srm).
[64] C. M. Reddy, A. Pearson, L. Xu, A. P McNichol, B. A. Benner,
Jr., S. A. Wise, G A. Klouda, L. A. Currie, and T. I. Eglinton,
Radiocarbon as a tool to apportion the sources of polycyclic
aromatic hydrocarbons and black carbon in environmental samples. Environ. Sci. Technol. 36, 1774-1782 (2002).
[65] L. A. Currie, G A. Klouda, B. A. Benner, Jr., K. Garrity, and
T. I. Eglinton, Isotopic and Molecular Fractionation in
Combustion; Three Routes to Molecular Marker Validation,
including Direct Molecular "Dating" (GC/AMS), Atm.
Environ. 33, 2789-2806 (1999).
[66] T. I. Eglinton, L. I. Aluwihare, J. E. Bauer, E. R. M. Druffel,
and A. P. McNichol, Gas chromatographic isolation of individual compounds from complex matrices for radiocarbon dating.
Anal. Chem. 68, 904-912 (1996).
[67] B. Hughey, P L. Skipper, R. E. Klinkowstein, R. E. Sheffer,
J. S. Wishnok, and S. R. Tannenbaum, Low-energy biomedical
GC-AMS system for ''^C and 'H detection, Nucl. Instrum.
Meth. Phys. Res. B172, 40-46 (2000).
[68] F. S. Rowland, Atmospheric Chemistry; Causes and Effects,
Marine Technol. Soc. J. 25, 12-18 (1991).

(2001), an invited NIST Sigma Xi lecture (1998), and a

plenary lecture at the Fourteenth Radiochemical
Conference (2002). At the Conference in 2002,
Dr. Currie was presented the l.M. Marci medal, the
highest award of the Czech Spectroscopic Society of the
Czech Academy of Sciences. The National Institute
of Standards and Technology is an agency of the
Technology Administration, U.S. Department of
Commerce.

About the author: Dr L.A. Currie is an NIST Fellow

Emeritus in the Chemical Science and Technology
Laboratory. The ideas behind this article were first
conceived about 15 years ago in connection with
lectures at NIH and the University of Maryland, and
as an outgrowth of the author's research on environmental radiocarbon while leader of the Atmospheric
Chemistry Group at NIST. The concept and scope of the
article were crystallized in connection with luncheon
talks at the Measurement Science Conference (1995)
and the Radiochemical Measurement Conference
217