1. Development of a dual MCP framing camera for high energy x-raysa)
N. Izumi, G. N. Hall, A. C. Carpenter, F. V. Allen, J. G. Cruz, B. Felker, D. Hargrove, J. Holder, J. D. Kilkenny, A.
Lumbard, R. Montesanti, N. E. Palmer, K. Piston, G. Stone, M. Thao, R. Vern, R. Zacharias, O. L. Landen, R.
Tommasini, D. K. Bradley, and P. M. Bell
Citation: Review of Scientific Instruments 85, 11D623 (2014); doi: 10.1063/1.4891712
View online: http://dx.doi.org/10.1063/1.4891712
View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/85/11?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Crosstalk in x-ray framing cameras: Effect on voltage, gain, and timing (invited)a)
Rev. Sci. Instrum. 83, 10E135 (2012); 10.1063/1.4740524
An x-ray streak camera with high spatio-temporal resolution
Appl. Phys. Lett. 91, 134102 (2007); 10.1063/1.2793191
Energy resolved fast two-dimensional x-ray imaging for MFE plasmas (invited)
Rev. Sci. Instrum. 75, 3926 (2004); 10.1063/1.1791750
Correcting for gain effects in an x-ray framing camera in a cylindrical implosion experiment
Rev. Sci. Instrum. 75, 3947 (2004); 10.1063/1.1787928
Further development of a single line of sight x-ray framing camera
Rev. Sci. Instrum. 74, 2191 (2003); 10.1063/1.1537850
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.115.190.41 On: Wed, 03 Aug
2016 19:55:24

2. REVIEW OF SCIENTIFIC INSTRUMENTS 85, 11D623 (2014)
Development of a dual MCP framing camera for high energy x-raysa)
N. Izumi,1,b)
G. N. Hall,1
A. C. Carpenter,1
F. V. Allen,1
J. G. Cruz,1
B. Felker,1
D. Hargrove,1
J. Holder,1
J. D. Kilkenny,2
A. Lumbard,1
R. Montesanti,1
N. E. Palmer,1
K. Piston,1
G. Stone,1
M. Thao,1
R. Vern,1
R. Zacharias,1
O. L. Landen,1
R. Tommasini,1
D. K. Bradley,1
and P. M. Bell1
1
Lawrence Livermore National Laboratory, Livermore, California 94550, USA
2
General Atomics, San Diego, California 92121, USA
(Presented 4 June 2014; received 2 June 2014; accepted 18 July 2014; published online 25 August
2014)
Recently developed diagnostic techniques at LLNL require recording backlit images of extremely
dense imploded plasmas using hard x-rays, and demand the detector to be sensitive to photons with
energies higher than 50 keV [R. Tommasini et al., Phys. Phys. Plasmas 18, 056309 (2011); G. N. Hall
et al., “AXIS: An instrument for imaging Compton radiographs using ARC on the NIF,” Rev. Sci.
Instrum. (these proceedings)]. To increase the sensitivity in the high energy region, we propose to use
a combination of two MCPs. The first MCP is operated in a low gain regime and works as a thick
photocathode, and the second MCP works as a high gain electron multiplier. We tested the concept of
this dual MCP configuration and succeeded in obtaining a detective quantum efficiency of 4.5% for
59 keV x-rays, 3 times larger than with a single plate of the thickness typically used in NIF framing
cameras. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4891712]
I. INTRODUCTION
X-ray framing cameras based on proximity-focused
micro-channel plates (MCP) have been playing an important
role as diagnostics of inertial confinement fusion experiments
for several decades.1–3
Most of the current x-ray framing cam-
eras consist of a single MCP, a phosphor, and a recording de-
vice (e.g., CCD or photographic films). This configuration is
successful for imaging x-rays with energies below 20 keV. In
this single MCP configuration, the MCP works as both the
photocathode and the electron multiplier. To have enough op-
tical output from the phosphor, the MCP has to be operated
with electron multiplication gains of 500–2000. When the
MCP is operated in this high gain regime, x-rays detected in
the shallow layer experience high electron-multiplication and
dominate the output, while x-rays detected deep inside have a
limited contribution. Due to this large gain differential, the de-
tective quantum efficiency (DQE) for volumetrically absorbed
photons (above 20 keV) is severely reduced.4
One approach
to mitigate this DQE reduction is to separate the photocathode
from the electron multiplier. We propose to stack two MCPs
and operate the first one in low gain regime, as a thick photo-
cathode, and use the second MCP as the electron multiplier,
to achieve enough optical output from the phosphor.
II. DQE AND NOISE FACTOR
Quantum efficiency (QE), which is defined as the num-
ber of detected events per number of incident quanta, is a
commonly used metric for estimating the statistical noise of
a)Contributed paper, published as part of the Proceedings of the 20th
Topical Conference on High-Temperature Plasma Diagnostics, Atlanta,
Georgia, USA, June 2014.
b)Author to whom correspondence should be addressed. Electronic mail:
izumi2@llnl.gov
the obtainable image when the output of the single detection
event has small variance. For detectors having broad pulse-
height distribution (PHD), DQE is the quantity of interest.5
The definition of the DQE is
DQE ≡
SNRo
SNRi
2
, (1)
where SNRo and SNRi are signal-to-noise ratio at output and
input of the detector, respectively. When the statistical be-
havior of the input photons follows Poisson statistics, the ex-
pected signal-to-noise ratio of the image can be evaluated as
SNRo = DQE × Ni, (2)
where Ni is the number of incident photons per resolution
element.
Another useful measure of the loss of the available infor-
mation caused by the statistical fluctuation of the avalanche
process is the noise factor (NF) defined as6
NF = 1 +
σ2
ξ 2
, (3)
where ξ is the mean and σ is the standard deviation of
the PHD. When noise from other sources (e.g., multiplica-
tive noise or statistical fluctuation of background exposure) is
small, the DQE can be expressed as7
DQE =
QE
NF
. (4)
III. EXPECTED PHD OF SINGLE MCP
The statistical uncertainty of an avalanche multiplica-
tion process is usually modeled by Pólya or Furry statistics.8
When an avalanche stream is started from a single electron
0034-6748/2014/85(11)/11D623/3/$30.00 © 2014 AIP Publishing LLC85, 11D623-1
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.115.190.41 On: Wed, 03 Aug
2016 19:55:24

3. 11D623-2 Izumi et al. Rev. Sci. Instrum. 85, 11D623 (2014)
and the number of collisions in the MCP pore is large, the
PHD can be approximated by a negative exponential,9
dP(ξ)
dξ avalanche
=
1
ξ
exp −
ξ
ξ
, (5)
where ξ and ξ are the gain and the average value of the gain.
Another source of PHD broadening is the depth depen-
dent gain. The electron multiplication for a detection event
that has taken place at a depth x from the MCP surface is
ξ(x) = ξ0
L−x
L , (6)
where ξ0 is the nominal gain of the surface event (x = 0),
and L is the thickness of the MCP. The PHD determined by
this depth dependence is10
dP(ξ)
dξ
=
L
0
dP
dx
×
dP(ξ, x)
dξ avalanche
dx
=
L
0
μ
1 − e−μL
exp(−μx)
1
ξ0
(1−x / L)
× exp −
ξ
ξ0
(1−x / L)
dx, (7)
where μ is the linear absorption coefficient of the MCP. For
hard x-rays (μL 1) this equation can be simplified:
dP(ξ)
dξ
=
1
ξ ln( ξ0 )
exp −
ξ
ξ0
− exp(−ξ) . (8)
Fig. 1 shows the PHD calculated by Eq. (7) (L: 460 μm, μ:
0.385 mm−1
at 59 keV). When ξ0 is high, the effect of the
depth dependent gain is dominant and the PHD shows a long
tail with ξ−1
dependence.
We evaluate the noise factor as
NF( ξ0 ) =
∞
0 ξ2 dP(ξ)
dξ
dξ
∞
0 ξ dP(ξ)
dξ
dξ
2
≈
ξ0 + 1
ξ0 − 1
ln( ξ0 ) (9)
and it has minimum = 2 when the MCP is operated near unity
gain: ξ0 ∼ 1. Fig. 2 shows a comparison of Eq. (9) and the
experiment. The estimated nominal gain (the first-order mo-
ment of Eq. (7)) is also plotted. To produce a strong enough
signal in a harsh neutron induced background environment,11
a MCP based x-ray detector has to be operated with the nomi-
nal gain about 103
( ξ0 ∼ 104
). Therefore, the expected noise
factor of cameras with a single MCP configuration is 8–9.
FIG. 1. Expected PHD calculated by Eq. (7).
FIG. 2. Comparison of the experimentally obtained noise factor and Eq. (9).
The surface gain of the data was estimated from a test sheet provided by the
manufacturer.
Therefore, the SNRo obtainable with the conventional single
MCP configuration is smaller by factor 3 than that of an ideal
detector which has a narrow PHD for high energy x-rays.
In order to obtain enough DQE with the nominal gain
∼103
, we propose to stack two MCPs. The first low-gain MCP
works as thick x-ray photocathode with low NF. The second
MCP amplifies the electron stream. To validate the advantages
of this concept, we assembled a test module and measured the
QE and the PHD of the single and dual MCP setup.
IV. EXPERIMENT
Fig. 3 shows the experimental setup we have used to
test the proposed configuration. The MCP was irradiated by
59 keV x-rays emitted by a 241
Am (10 μCi) radioactive
source, 48.5 mm from the surface of the MCP. The low energy
emission lines from the source (13.9, 26.3, 33.2, and
43.4 keV) were filtered by an aluminum filter (3.88 mm thick)
which also acted as the vacuum window. X-ray photons are
converted to electrons in the MCPs. After multiplication in the
MCPs, the electrons are accelerated to 3 keV and hit the P46
phosphor (Y3Al5O12:Ce). The optical output from the phos-
phor was detected by the photo-multiplier tube (PMT: Hama-
matsu R329-02, bias: 1.1 kV).
The PMT signal was amplified by a charge sensitive am-
plifier (ORTEC 113), shaped (ORTEC 460) and recorded by a
multi-channel analyzer (MCA: ORTEC TRUM PCI 8K). The
dynamic range of the pulse height recording system is limited
to about 100. To extend the dynamic range, we performed the
measurements twice with different gain setting (0.15–40 pC
and 1.2–400 pC range) and spliced the respective spectra at
20 pC. Background counts from the PMT, the MCP, and the
Am 241 MCP
Low gain
Phosphor / FOP
Al filter
MCP
High gain
PMT
FIG. 3. Experimental setup.
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.115.190.41 On: Wed, 03 Aug
2016 19:55:24

4. 11D623-3 Izumi et al. Rev. Sci. Instrum. 85, 11D623 (2014)
FIG. 4. Experimentally obtained pulse height distribution: (a) single MCP
460 μm thick and (b) dual MCP 460 μm + 460 μm.
direct x-ray excitation on phosphor were measured separately
and subtracted from the raw data.
Fig. 4(a) shows the PHD obtained with single MCP
(thickness L: 460 μm, pore diameter d: 10 μm). When the
MCP is operated in low gain, the PHD is close to a decaying
exponential and the standard deviation of the PHD is close to
the mean value. When we increase the MCP gain by applying
higher bias voltage, the depth dependent gain effect becomes
significant. We also measured the PHD of a 800 μm thick sin-
gle MCP. Fig. 5(a) shows the QE and DQE of the single MCP
configuration as a function of the bias voltage. The observed
DQE departs from the QE behavior at high gain values due
to the depth dependent gain effect. When the nominal MCP
gain is 103
(the nominal PMT signal ∼18 pC/event), the ob-
served DQE was 1.4% (2.6%) for 460 μm (800 μm) thick
plate.
Fig. 4(b) shows the result of the dual plate configuration
(stack of two L/d = 46 plates). We changed the bias volt-
age given to the 1st MCP (V1) from 0 to 570 V, while the
2nd MCP (V2) was operated with 600 V bias. The voltage
given to the gap between two plates (Vg) was 100 V. When V1
= 0 V, the events are dominated by x-ray detection on the 2nd
MCP and the observed PHD was almost identical to that of the
single plate with 600 V bias. When the V1 ≥ 300 V, increase
of the count rate was observed below 0.3 pC. We believe this
is false count due to multiple counting of the large x-ray de-
tection events caused by slow decay component of the P46
phosphor. Therefore, the PHD below 0.3 pC was ignored for
calculation of the QE and the DQE. When the 1st plate was
operated near the turn on voltage (V1 ∼ 400 V), the PHD in-
creased starting from a few pC without affecting the high gain
tail. When the gain was set higher than unity, the PHD started
FIG. 5. Measured QE and DQE of the single MCP: (a) single MCP and
(b) dual MCP.
showing ξ−1
dependence. Fig. 5(b) shows the QE and DQE
for the dual MCP configuration. The DQEs were compared
for 1st MCP thicknesses of 460 μm and 800 μm (L/d: 46 and
80, respectively). The highest DQE (4.5%) is obtained when
the 800-μm-thick plate with V1 = 625 V. When the nominal
MCP gain is 103
, the observed DQE is 3.2% and 4.3% for the
460-μm-thick and 800-μm-thick, 1st MCP, respectively.
V. DISCUSSION
By using the dual MCP configuration, the noise factor
of the system was reduced and the DQE was successfully in-
creased by a factor ∼3 compared to commonly used 460 μm
thick single MCP configuration. However, reduced QE was
experienced when the gain of the 1st MCP was set to low val-
ues. We believe this is due to the termination of the electron
stream on the first collisions with the pore wall (causing a zero
multiplication). If that is the case, the QE may be improved by
reducing the L/d ratio of the 1st MCP because it is possible to
accelerate the electrons to a higher energy per collision while
keeping low overall multiplication gain.6
ACKNOWLEDGMENTS
This work was performed under the auspices of the U.S.
Department of Energy by Lawrence Livermore National Lab-
oratory under Contract No. DE-AC52-07NA27344.
1J. D. Kilkenny, Laser Part. Beams 9, 49 (1991).
2R. Tommasini et al., Phys. Plasmas 18, 056309 (2011).
3G. N. Hall et al., “AXIS: An instrument for imaging Compton radiographs
using ARC on the NIF,” Rev. Sci. Instrum. (these proceedings).
4J. E. Bateman et al., Nucl. Instrum. Methods 144, 537 (1977).
5G. Zanella and R. Zannoni, Nucl. Instrum. Methods Phys. Res. A 359, 474
(1995).
6A. J. Guest, Acta Electron. 14, 49 (1971).
7G. Zanella and R. Zannoni, Nucl. Instrum Methods Phys. Res. A 406, 93
(1998).
8S. E. Moran et al., Proc. SPIE 3173, 430–457 (1997).
9J. L. Wiza, Nucl. Instrum. Methods 162, 587 (1979).
10K. W. Dolan and J. Chang, Proc. SPIE 0106, 178–188 (1977).
11N. Izumi et al., Rev. Sci. Instrum. 81, 10E515 (2010); C. Hagmann et al.,
ibid. 81, 10E514 (2010).
Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 128.115.190.41 On: Wed, 03 Aug
2016 19:55:24