Applied Radiation and Isotopes 109 (2016) 193197

Contents lists available at ScienceDirect

Applied Radiation and Isotopes

journal homepage: www.elsevier.com/locate/apradiso

Performance of an in-situ alpha spectrometer

R. Pllnen
STUK-Radiation and Nuclear Safety Authority, P.O. Box 14, FI-00881 Helsinki, Finland

H I G H L I G H T S

Novel equipment for in-situ alpha spectrometry allows to detect also beta particles.
Use of a collimator enables good energy resolution for the alpha particles.
Alpha-particle detection efciency for wide area sources is 0.14 without collimation.
Alpha-particle detection efciency is by a factor of ten smaller with collimation.
Collimation provides no advantages for the beta particle detection.

art ic l e i nf o

a b s t r a c t

Article history:
Received 8 April 2015
Accepted 24 November 2015
Available online 2 December 2015

Equipment was recently developed for detecting alpha particles from at and smooth surfaces with good
energy resolution at ambient air pressure. In this work, the detection efciencies were determined using
different extended-area sources emitting alpha and beta radiation and a mixed nuclide point source
emitting alpha radiation. Beta particles are of importance because they can also be detected. Counts
originating from alpha and beta particles are mainly at different energies, which make their separation
possible. An efciency of 0.14 was determined for an extended-area (4 30 cm2) homogeneous source
emitting alpha radiation at the energy of 56 MeV, whereas for the beta emitters the efciencies were
0.070.19 depending on the beta-particle emission energies. The use of a collimator reduces the detection efciencies by a factor of up to ten.
& 2015 Elsevier Ltd. All rights reserved.

Keywords:
In-situ alpha spectrometry
Silicon detector
Detection efciency

1. Introduction
A novel approach and equipment ADONIS for in-situ alpha
spectrometry were recently introduced (Pllnen et al., 2012,
2015). No vacuum is necessary for particle detection, which means
that alpha spectrometry can be performed in the eld or indoors
using handheld instruments similar to those used in gamma-ray
spectrometry. In the present paper performance of the equipment
in terms of the detection efciency is presented.
Alpha spectra with good energy resolution can be obtained
from at and smooth surfaces using a collimator and a semiconductor silicon detector operating at ambient air pressure. A
honeycomb collimator placed in front of the active volume of the
detector allows detecting alpha particles entering perpendicularly
to the detector. In xed measurement geometry the alpha-particle
path length in air is almost constant, thereby making good energy
resolution possible. However, using the collimator may cause notably lower detection efciency.
Preliminary tests of the equipment were reported earlier
E-mail address: roy.pollanen@stuk.
http://dx.doi.org/10.1016/j.apradiso.2015.11.110
0969-8043/& 2015 Elsevier Ltd. All rights reserved.

(Pllnen et al., 2015). In the present paper the results of detection

efciency tests are presented using well-characterized sources
emitting alpha and beta radiation. Beta particles, electrons and
low-energy photons are of special importance because they may
produce counts at the low-energy part of the energy spectrum. To
some extent, they may interfere with the signals originating from
alpha particles and thus deteriorate the alpha spectrum analysis.
On the other hand, the fact that beta-particle emitting radionuclides may be detected as well widens the application area of
the equipment. Detection efciencies with and without the collimator are determined here for extended-area sources and a point
source emitting alpha radiation.
A special feature of the measurement system is the possibility
of using remote expert support (reachback) for the data visualization and analysis. Certain tasks, such as spectrum unfolding, can
be done more easily in a remote ofce than in harsh environmental conditions. In addition, the reachback concept enables
obtaining more reliable results when no experienced personnel is
available for the eld measurements. In this paper, the use of the
reachback concept is briey illustrated.

194

R. Pllnen / Applied Radiation and Isotopes 109 (2016) 193197

Table 1
Characteristics of the ve extended-area (10  10 cm2) sources containing one radionuclide and one mixed nuclide point source. Notations t1/2, E and E,max refer to the
half life, emission energy of the alpha particles, and end point energy of the beta particles, respectively. Data are rounded and only emissions with the largest yields
(percentages in the parentheses) are mentioned. Full data with uncertainties can be obtained from http://www.nucleide.org/DDEP_WG/DDEPdata.htm. A and Aa refer to the
activity and activity per unit area, respectively. Uncertainties for the activities and activities per unit area only refer to statistical uncertainty (k 2) and do not include
systematic errors. Data acquisition times are without (tacq) and with (tacq,coll) collimation. In the 90Sr source the short-lived daughter 90Y was in secular equilibrium with its
parent. In the 239Pu source there was small amount of 241Am (0.020 (3) Bq cm  2, k 2) generated by the decay of 241Pu.
Extended-area sources
Nuclide

t1/2

E (keV)

E,max (keV)

Aa (Bq cm  2)

tacq (h)

tacq,coll (h)

241

433 a
2.41  104 a
5.27 a
3.02  105 a
28.8a 2.67 d

5486 (84%)
5157 (71%)

317 (100%)
710 (98%)
546 (100%) 2280 (100%)

11.8 (4)
10.4 (3)
6.10 (31)
8.36 (42)
11.1 (6) 11.1 (6)

8.5
10.4
10.9
8.7
11.7

12.6
11.6
11.3
9.6
10.5

E,max

A (Bq)

tacq (h)

tacq,coll (h)

5486 (84%)
5157 (71%)
5805 (77%)

376 (13)
431 (15)
304 (14)

7.3
7.3
7.3
30.8

9.5
9.5
9.5
52.2

Am
Pu
60
Co
36
Cl
90
Sr90Y
239

Point (mixed nuclide) source

Nuclide
t1/2
241

Am
Pu
Cm
Background
239

244

433 a
2.41  104 a
18.11 a

Fig. 1. Count rates from the extended-area sources (a) 60Co, (b) 36Cl, (c) 90Sr90Y, (d) 239Pu and the background. Topmost spectra (solid line) were measured without
collimation whereas spectra in the middle (dotted line) were measured with the collimator. Background spectra (at the bottom) measured without and with collimation
were practically identical. Channel number 150 is taken as a borderline between counts originating from alpha and beta particles.

2. Equipment and sources

The operation principle and characteristics of ADONIS were
reported previously (Pllnen et al., 2012,2015) and are only
briey repeated here. The alpha detector used in the equipment

(CAM2000AM, Canberra) was mounted in the detector head in

addition to the preamplier, collimator and Mylar foil protecting
the detector. The cell size of the honeycomb collimator is 6.3 mm
and the cell wall thickness is 64 mm. The source to detector distance was xed at 12 mm. The area from which the alpha particles

R. Pllnen / Applied Radiation and Isotopes 109 (2016) 193197

Fig. 2. Count rate per Bq cm  2 for extended-area sources containing

90
Sr90Y and 239Pu measured without collimation.

36

Cl,

195

60

Co,

can enter the detector is 29.5 cm2, corresponding to a diameter of

61.3 mm. Although the detector is primarily designed for alpha
particle detection, it can also be used for the detection of betaparticle emitting radionuclides. However, in this case only a fraction of the beta particle energy may be detected because of the
small active thickness (300 mm) of the detector. The number of
channels in the spectra was 1024. The signal amplication settings
are such that beta particles generate counts mainly below channel
150, whereas signals from alpha particles are usually at higher
channel numbers.
In the case of homogeneous contamination, extended-area
sources with activities of approximately 10 Bq cm  2 (Table 1)
were used to investigate the detector response. For these sources
the geometrical detection efciency calculated by the AASI program (Siiskonen and Pllnen, 2005) was geom 0.185 (0.1% relative uncertainty in the computation). A mixed 239Pu, 241Am and
244
Cm alpha particle source with a sum activity of approximately
1 kBq (Table 1) was used to determine the detection efciency for
a point source. Although the diameter of the active area was 5 mm,
it can be considered as a point source from the perspective of the
detector. The calculated geometrical point source efciency was
geom 0.285, assuming that the source is located on the symmetry axis of the detector (0.1% relative uncertainty). All sources
were measured with and without the collimator.

3. Determination of detection efciencies

Determination of the detection efciency is necessary to relate
the number of detected counts to the activity per unit area. To
perform this, counts originating from the beta particles must be

Fig. 3. Count rates measured from the mixed nuclide point source without and
with collimation. With collimation the Full Width at Half Maximum (FWHM) of the
239
Pu, 241Am and 244Cm peaks was 130140 keV. The simulated spectrum (dotted
line) is practically identical to that measured without collimation. The activity of
the source is high enough to generate random coincidence counts above channel
500 when the measurement was done without collimation.

distinguished from those caused by the alpha particles. Beta particles generate counts mainly below channel 150 (see Figs. 1 and
2), whereas the counts observed at higher channels originate from
alpha particles. This channel number corresponds to an alphaparticle emission energy of approximately 3.9 MeV (Pllnen et al.,
2015). Radionuclides emitting high-energy beta particles, such as
90
Y (Fig. 1c), may generate signals above channel 150, but their
contribution to the total number of counts is small (10  4). Determination of the detection efciency below channel 150 is
questionable if the source emits both beta and alpha particles. For
the present measurements, the total background count rate was
about 1 s  1, but the background count rate above channel 150 was
only 2  10  3 s  1.
Detection efciencies were calculated separately for alpha
(channels 4 150) and beta particles (channels r150) emitted
from the extended-area sources (Table 2). The results obtained
from the extended-area 241Am source were reported in Pllnen
et al. (2015) and are presented in Table 2 for comparison. The efciencies were calculated from the ratio of detected (and background-corrected) and emitted particles.
When using collimation, the alpha-particle detection efciencies dropped by a factor of about ten. For the high-energy beta
emitter 90Sr90Y, the collimator reduced the efciency by a factor
of two, whereas for the low-energy beta emitter 60Co the reduction factor was 3.6. It can be concluded that high-energy beta
particles can penetrate the collimator walls. The alpha detector can
also detect signals from high-energy gamma-ray interactions with

Table 2
Detection efciencies for the extended-area sources. Uncertainties refer to the statistics (coverage factor k 2, see main text) and do not take into account systematic errors.
For the alpha emitters, the efciencies were calculated based on channels 4150 whereas channels r150 were used for the beta emitters. The activity per unit area needed
to produce one count per second is denoted as 1 cps-Bq cm  2, whereas notation 1 Bq cm  2-cps gives the number of counts per second that correspond to the unit
activity per area.
Nuclide

No collimation

241

Am
Pu
Co
36
Cl
90
Sr90Y
239
60

0.143
0.128
0.068
0.188
0.192

(5)
(4)
(3)
(9)
(10)

Collimation
1 cps-Bq cm  2

1 Bq cm  2-cps

Coll

0.23
0.25
0.50
0.18
0.18

4.2
4.0
2.0
5.4
5.7

0.016
0.016
0.019
0.061
0.100

(1)
(1)
(1)
(3)
(5)

1 cps-Bq cm  2

1 Bq cm  2-cps

2.1
2.0
1.8
0.56
0.34

0.47
0.49
0.56
1.8
2.9

196

R. Pllnen / Applied Radiation and Isotopes 109 (2016) 193197

Fig. 4. Snapshot from one of the reachback screens. Waterfall representation of the spectra obtained from different test sources emitting alpha and beta radiation is
presented in the top left window (vertical and horizontal axes refer to time and energy, respectively). From the waterfall picture the user can select appropriate spectra for
standard spectrum presentation (window in the top right, vertical axis refers to the number of counts per channel). These spectra can be (automatically) analyzed using e.g.
the ADAM program (Ihantola et al., 2011). Location of the measurements can be viewed on the map (window in the bottom, the location option was not used in the
abovementioned test measurements).

materials surrounding the active volume of the detector.

The alpha detection efciencies for the mixed nuclide point
source located on the symmetry axis of the detector were 0.224
(8) and 0.023 (5) (coverage factor k 2), when the measurements
were done without and with collimation, respectively (Fig. 3).
These numbers were approximately 50% higher than in the case of
homogeneous contamination. In the uncollimated case, and if all
the detected counts were taken into account (i.e. in the absence of
beta emitters), the detection efciency was 0.261 (9), which is
practically equal to the geometrical detection efciency.
Simulations for the ADONIS measurement geometry were
performed using the AASI program to compute the number of
totally absorbed alpha particles. These are the particles that would
enter the active volume of the detector without the presence of
absorbing material (mainly air). The shape of the simulated
spectrum (dotted line in Fig. 3) was almost identical to the measured spectrum. In the point-source measurement geometry the
contribution of the totally absorbed alpha particles was 8% from
those emitted towards the detector. In the measured spectrum,
14% of the alpha particles emitted towards the detector were detected in channels r150.
The point source measurements without collimation can be
expected to have a lower efciency if the source is located off the
symmetry axis. In general, the point source activity estimation
cannot be done without collimation if the source location is unknown. However, when using the collimator the source location
need not to be known, because every honeycomb cell acts like an
individual detector. The relationship between the activities of the
source to the detected number of counts can be determined as

described above. For a point source one count per second refers to
43.3 Bq and one Bq refers to 0.023 counts per second.
The uncertainties reported in Table 2 refer to the uncertainties
in the source activities and the statistics. However, in the measurement and data analysis procedures there are systematic uncertainties that must be taken into account but quantitative estimation of their detailed contribution is beyond the scope of the
present paper. The selection of the borderline channel 150 for
separating alpha particle counts to those generated by beta particles is an example of a systematic uncertainty. Larger alpha detection efciencies may be obtained by selecting smaller value for
the borderline channel but at the same time some beta-particle
counts may be erroneously considered as counts originating from
alpha particles. In addition, the non-collimated alpha-particle detection efciencies presented in Table 2 are valid for nuclides
whose emission energies are approximately 56 MeV. The efciencies will be somewhat larger for nuclides with higher emission
energies (but not larger than the geometrical detection efciency
of 0.185), whereas the efciencies may be drastically lower for
low-energy alpha emitters. This is because the contribution of
counts below channel 150 is considerably larger for low-energy
emitters than for high-energy emitters. Another notable systematic uncertainty is due to backscattering, especially in the case
of beta particles. In the present case the source materials are on an
aluminum disk and the beta particle surface emission rates could
be up to 30% higher than for a non-scattering backing material.
This is the reason why the detection efciency of 90Sr90Y was
found to be larger than the geometrical detection efciency.
Conversions between total count rates and activities per unit

R. Pllnen / Applied Radiation and Isotopes 109 (2016) 193197

area (Table 2) are of particular importance especially when the

personnel operating in the eld are not familiar with spectrum
analysis. They facilitate operation and situation awareness in the
eld as well as in a reachback center, which provides expert
support for the eld teams.

4. Remote expert support

The data acquisition program of ADONIS automatically sends
the measurement data to a remote 'Linssi' SQL database for gamma-ray spectrometry (Aarnio et al., 2008). The possibility of
making spectrum analysis in the eld is not ruled out, but thorough data investigation may be inconvenient in the eld due to
the circumstances or possibly the inexperience of the eld operator. In addition, all relevant information is not necessarily accessible for the eld teams. Therefore it is an asset to have the
option to send the measurement and other relevant data to a remote location in which expert support is available. Remote expert
support is an essential functionality of the equipment. Experts
operating in the reachback can easily select different views, such
as counts-per-second view or waterfall view, to visualize the data
(Fig. 4).

Discussion
In-situ alpha and beta spectrometry with silicon detectors may
be done in a laboratory, where the sources can be radiochemically
processed and where the measurements can be performed in a
stable environment. However, in some cases rapid measurements
from non-optimal sources in uncontrolled environmental conditions may be necessary, which pose challenges for the equipment,
measurement procedures, data transfer and spectrum analysis.
Although viable solutions exist for all these areas, there is still
room for improvement. For example, performing correct energy

197

calibration is an intractable problem because of the nonlinearities

and differences in the energy loss of alpha and beta particles in air.
Another open area is the development of spectrum unfolding
software to take into account complex peak shapes arising from
non-ideal sources. Nevertheless, as shown in the present paper,
user-friendly in-situ measurements are possible provided that
measurement procedures, data transfer and processing, and procedures for spectrum analysis are designed properly.

Acknowledgments
This work was performed within the EMRP Research Programme under the JRP-Contract IND57 MetroNORM. The AASI
program can be downloaded free of charge from website https://
www.stuk./web/en/services/aasi-program-for-simulating-en
ergy-spectra-in-alpha-spectrometry. Mr. Samu Ristkari and Ms.
Tarja Ilander are kindly acknowledged for the reachback support
and providing Fig. 4.

References
Aarnio, P., Ala-Heikkil, J., Isolankila, A., Kuusi, A., Nikkinen, M., Siiskonen, T., Toivonen, H., Ungar, K., Zhang, W., 2008. LINSSI: Database for gamma-ray spectrometry. J. Radioanal. Nucl. Chem. 276, 631637.
Ihantola, S., Pelikan, A., Pllnen, R., Toivonen, H., 2011. Advanced alpha spectrum
analysis based on the tting and covariance analysis of dependent variables.
Nucl. Instrum. Methods A656, 5560.
Pllnen, R., Perjrvi, K., Siiskonen, T., Turunen, J., 2012. High-resolution alpha
spectrometry at ambient air pressure Towards new applications. Nucl. Instrum. Methods A694, 173178.
Pllnen, R., Turunen, J., Karhunen, T., Perjrvi, K., Siiskonen, T., Wirta, M., Turunen, A., 2015. Novel equipment for in-situ alpha spectrometry with good
energy resolution. Health Phys. 109, 601605.
Siiskonen, T., Pllnen, R., 2005. Advanced simulation code for alpha spectrometry.
Nucl. Instrum. Methods A550, 425434.