solar activity and ozone layer depletion - Inter Islamic Network on ...

SOLAR ACTIVITY AND OZONE LAYER DEPLETION
COMPARISION OF PARAMETRIC
EVALUATION OF TOTAL OZONE MAPPING
SPECTROPHOTOMETER (TOMS) DATA
NIMBUS -7 WITH THE DOBSON SPECTROPHOTOMETER
OBSERVATIONS AT
THE GEOPHYSICAL CENTER QUETTA, PAKISTAN
M.Ayub Khan Yousuf Zai, Jawaid Quamar,M.Rashid Kamal Ansari,
Jawaid
Iqbal <strong>and</strong> Arif Hussian
Institute of Space <strong>and</strong> Planetary Astrophysics <strong>and</strong> Department of Applied
Physics, University of Karachi
Abstract
In this communication we present some approaches to relate the <strong>depletion</strong>
process with the <strong>solar</strong> <strong>activity</strong>. We propose to study the effectiveness of <strong>ozone</strong>
<strong>layer</strong> structure. Developing a quantitative treatment for data covering a specific
period, this study will report for Pakistan an explorative assessment of the link
between <strong>ozone</strong> <strong>layer</strong> <strong>depletion</strong> <strong>and</strong> <strong>solar</strong> <strong>activity</strong>. We have computed
parameters, which signify the measurements drawn from the satellite, <strong>and</strong>
ground based instruments. We will also try to look for underst<strong>and</strong>ing the
interaction of dynamics <strong>and</strong> <strong>ozone</strong> photochemistry in the middle atmosphere (
25-35 km) <strong>and</strong> the nature of stratospheric responses to anomalous <strong>solar</strong> <strong>activity</strong>.
We have compared both the data sets using different techniques <strong>and</strong> observed
that the trend follows the same path. This could be helpful in obtaining the
estimates of sunspots <strong>and</strong> establishing a correlation with the <strong>ozone</strong> <strong>layer</strong>
reduction then suitable predictions can be made for our government <strong>and</strong> private
departments
Introduction
Solar studies are of importance in general astronomy because the sun is the only
star upon which surface detail an be resolved The investigators find the variety of
<strong>solar</strong> structures <strong>and</strong> events to decide their physical causes. For our society, the
study of the sun is more indirect but of great importance because <strong>solar</strong> radiation
is essential to our biosphere. Cyclicity of <strong>solar</strong> <strong>activity</strong> (<strong>solar</strong> cycle) could be a
factor supporting the emergence of OLD (Solar terrestrial relations analysis). It is
suspected that Cl2 is produced during active period of the sun <strong>and</strong> this production
of Cl2 causes Ozone <strong>layer</strong> <strong>depletion</strong>.
The rates of <strong>ozone</strong> removal in the year 1992 are enhanced by volcanic eruptions
in 1991 <strong>and</strong> 1992. Because anthropogenically introduced chlorine(Cl) <strong>and</strong>
bromine(Br) levels will remain high for so long, it is expected that there will be

more <strong>and</strong> more UV-B reaching the earth’s surface. Man-made causes of OLD
are providing a threshold for OLD. We can not neglect the natural events which
are occurring in the upper atmosphere <strong>and</strong> producing certain chemical
constituents which react with O3 <strong>and</strong> dissociate it into nascent oxygen <strong>and</strong> a
molecular oxygen an can be seen by Chapman reactions. Therefore, in view of
this conception the Montreal Protocol is not to be all of the things.[1-8]. These
investigations led to series of processes known as Chapman mechanism which
explained the formation <strong>and</strong> annihilation of O3 <strong>layer</strong>. Using this mechanism,
predictions were made which provided excellent agreement with observations
available then [30-35]. The situation can be explained by the following reactions:
O2 + h ν → O + O
(1)
O + O2 + M → O3 + M
(2)
O3 + hν → O + O2
(3)
O + O3 → O2 + O2
(4)
M being a third body required carrying off the excess energy of the association
process. The rapid cycle composed of reaction (1.2) <strong>and</strong> (1.3) does not, of
course, destroy O3. Many years after the preceding initial work, it was suggested
that catalytic processes can lead to the overall reaction as follows:
(5)
O + O3 → 2 O2
Ozone content in the stratosphere varies due to various natural <strong>and</strong>
anthropogenic activities. Some prominent such activities creating OLD are
described as follows
(a) Technological
CFCs produced in various technological processes gradually migrate upward into
the stratosphere. At altitude above 30 km, intense UV radiation breaks down the
CFCs, causing them to release chlorine <strong>and</strong> destroying <strong>ozone</strong> as shown in the
following chemical reactions:
CFCl3 + h ν → CF Cl2 + Cl −
(6)
CF2Cl2 + h ν→ CF2 Cl + Cl
(7)

Cl + O3 → Cl O + O2
(8)
During the past few years, satellite <strong>and</strong> ground-based observations have
particularly contributed to our underst<strong>and</strong>ing of the general features of the
vertical <strong>ozone</strong> distribution <strong>and</strong> latitude variations in the stratosphere <strong>and</strong> lower
mesosphere. Ozone photochemistry in the middle atmosphere (25-35 km)
will contribute to link stratospheric responses to anomalous <strong>solar</strong> <strong>activity</strong>
(b) Solar Cycle
In the stratosphere sun drives the photochemistry as well as providing the
heating that drives the <strong>solar</strong> circulation. Because the atmosphere absorbs UV
<strong>and</strong> shorter wavelength radiation, accurate determination of its time variation
should be measured from above the top of the atmosphere. Some information
about <strong>solar</strong> variability can be obtained by consideration of longer wavelength
features that do penetrate to the ground, as the same mechanisms that give rise
to the longer-wavelength variations the most often is the radio radiation with a
10.7-micron wavelength, albeit the other proxies such as sunspot numbers can
also be used. Two of the most important periods for <strong>solar</strong> variations are 27 days
<strong>and</strong> 11 years.[9-22]
The <strong>solar</strong> cycle dependence of atmospheric <strong>ozone</strong> also has an altitude
dependence which maximises at approximately 45 km. The temperature changes
of the upper stratsphere are rather caused by the <strong>solar</strong> cycle variation. Basis for
the <strong>solar</strong> cycle theory was the notion that reactive nitrogen compounds (NO <strong>and</strong>
NO2), like chlorine can destroy stratospheric <strong>ozone</strong> quite effectively. These
compounds are produced in abundance in the mesosphere <strong>and</strong> lower
thermosphere (∼ 60-70 km) during period of high <strong>solar</strong> <strong>activity</strong>. The nitrogen
compounds are rapidly destroyed in the mesosphere, but might be transported all
to the stratosphere in the dark polar winter.
(c) Solar Proton Events
Sometimes the sun emits large amount of protons which are channeled by the
earth’s magnetic field to impact the earth at high altitudes. These are referred to
as <strong>solar</strong> proton events (SPEs). For protons of sufficient energy, they can reach
the mesosphere <strong>and</strong> stratosphere <strong>and</strong> lead to reduction of <strong>ozone</strong> by producing
increased concentrations of odd nitrogen <strong>and</strong> / or odd hydrogen compounds in
the stratosphere. A principal advance in the determination of <strong>ozone</strong>
distributions <strong>and</strong> variability came with the advent of satellite-based systems.
The most important of these are given as follows:
(1) Total Ozone Mapping Spectrometer (TOMS) instruments for total <strong>ozone</strong>
measurements
(2) Solar Backscatter Ultraviolet (SBUV,I <strong>and</strong> II )
(3) Stratosphere Aerosol <strong>and</strong> Gas Experiment (SAGE, I <strong>and</strong>II) instruments for
vertical <strong>ozone</strong> distributions

(d) Investigating the correlations between the UV-B due to OLD <strong>and</strong> the
<strong>solar</strong> variations using modeling strategies
Ozone Layer depths (Dobson Units)
420
380
340
300
260
220
sunspot
y=279.266+0.051*x+eps
180
-20 20 60 100 140 180 220
Sun spot numbers (Counts)
Figure.1. Correlation between <strong>ozone</strong> concentration <strong>and</strong> the sunspot numbers
Measurements from satellite can give near-global covering, however, it provides
only averaged values of the measured quantity over large distances both in
vertical <strong>and</strong> horizontal directions. So we have preferably utilised Dobson data for
our computations <strong>and</strong> compared these with the data. down loaded from satellite
(TOMS, Nimbus-7 ). As far as the trend <strong>and</strong> cumulative behaviour of TOMS
satellite data <strong>and</strong> Dobson data measured with the help of ODD
spectrophotometer that was installed at Geophysical Center Quetta for the
duration 1979-1993 are concerned we have tried to demonstrate through precise
computations the comparison between these two data sets.
<strong>Inter</strong>national Ozone <strong>Network</strong> presently utilizes <strong>ozone</strong> depth detecting
spectrophotometer (ODDS). On board the satellite, incident <strong>solar</strong> radiation <strong>and</strong>
backscattered UV sunlight are measured. Total <strong>ozone</strong> is derived from these
measurements. To map total <strong>ozone</strong>, total <strong>ozone</strong> mapping spectrometer (TOMS)
instrument scan through the sub-satellite point in a direction perpendicular to the
orbit plane [18][23]. Details of these instruments could be found in every <strong>ozone</strong>
detection techniques literature.
Satellite-borne instruments (e.g. on Nimbus-7, Meteor-3, ADEOS <strong>and</strong> Earth’s
probe) are used to measure decreases in the concentration of stratospheric

Parameters
/ Years
<strong>ozone</strong> in our atmosphere. Total Ozone Mapping Spectrometer (TOMS) data
(Nimbus-7) are available for the period from 1979-1993.
A comparison TOMS <strong>and</strong> Dobson values using coefficient of variation (CV) for
each year of both the data sets, observing trends <strong>and</strong> calculating some other
parameters is given in Table.1.
TABLE.1: Comparison of parametric computation for TOMS satellite <strong>and</strong>
Dobson data during the session 1979-1993 <strong>and</strong> Dobson data
during the session 1979-1993
Mean of TOMS
/Dobson
X
TR – Mean
TOMS /Dobson
TR – X
Sta. Dev.
TOMS
/Dobson
σ
SE mean
TOMS
/Dobson
SE- X
Coefficient of
Variation
TOMS
/Dobson
(CV)
1979 285.76/ 298.25 284.18/296.32 23.01/21.79 1.30 / 6.29 0.08 / 0.07
1980 277.50/ 294.00 276.37/293.23 18.41/11.27 0.98 / 3.25 0.07 / 0.04
1981 280.89/291.50 279.62/ 290.45 16.85/13.32 0.88 / 3.84 0.06 / 0.05
1982 287.07/301.25 285.28/299.13 24.85/20.46 1.31/5.91 0.08 / 0.07
1983 272.55/283.92 271.64/281.34 18.45/8.93 0.97 / 2.56 0.07 / 0.03
1984 275.88/ 287.67 275.83/285.41 21.81/17.07 1.14/4.93 0.08 / 0.06
1985 262.09/ 277.42 262.15/276.21 12.71/ 6.68 0.67 / 1.93 0.05/ 0.02
1986 272.01/ 292.58 271.43/ 290.65 18.52/13.81 0.97 / 3.98 0.07 / 0.05
1987 270.81/ 286.92 271.12/ 285.22 23.12/20.64 1.21 / 5.96 0.08 / 0.07
1988 264.50/281.83 264.31/280.34 19.70/ 7.93 1.03/2.28 0.07 / 0.03
1989 279.62/297.75 279.13/280.34 15.43/10.57 0.81 / 3.05 0.05 / 0.04
1990 274.46/282.83 273.87/280.32 16.73/27.07 0.83 / 7.80 0.06 / 0.09
1991 280.25/297.83 280.23/295.11 20.52/15.39 1.08 / 4.44 0.07 / 0.05
1992 272.70/278.33 271.45/ 277.76 23.27/18.49 1.22 / 5.33 0.08 / 0.07
1993 262.54/264.92 262.77/ 263.22 18.10/5.82 1.61 / 1.68 0.06 / 0.02
TABLE. 2: Trend equations
1 TOMS-Nimbus – 7 Yt = 281.37 0 .85 t + ε
2 Dobson Instrument Yt = 297.23 1.179 t + ε
This justifies our adoption of Dobson data for computation <strong>and</strong> modeling. Data
having less variation is more stable <strong>and</strong> vice versa. Coefficient of variation hence
determines the stability or consistency of the data.

Ozone depth in DU for Quetta Pakistan
305
300
295
290
285
280
275
270
265
260
260
0 2 4 6 8 10 12 14 16
Time in years (1979-1993)
290
284
278
272
266
Ozone depth in DU from TOMS Nimbus-7
NEWVAR18 (L)
NEWVAR19 (R)
Figure. 2. Manifestation of the trends of TOMS <strong>and</strong> Dobson data sets
Ozone depth from Dobson
305
300
295
290
285
280
275
270
265
260
260
0 2 4 6 8 10 12 14 16
Time in years (1979-1993)
290
284
278
272
266
Ozone depth from TOMS Nimbus-7
DOBSON (L)
TOMS (R)
Figure 3: Depiction of the trends of TOMS <strong>and</strong> Dobson data sets with 95 %
confidence level.

Total Ozone from TOMS Nimbus-7
305
300
295
290
285
280
275
270
265
260
260
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Time in years (1979-1993)
290
284
278
272
266
Total Ozone from Dobson
DOBSON (L)
TOMS (R)
Figure. 4: Comparison of ground-based data recorded at Quetta Pakistan
on Dobson Spectrophotometer with the
TOMS Nimbus-7
The trend equations are also evolved using least square methods as shown in
table 2. These equations could be used for future predictions for both the data
sets. Accuracy of the Dobson Unit (DU) is based on the laboratory
measurements of the <strong>ozone</strong> absorption coefficients as mentioned above. An
error in the absorption coefficients is due to the b<strong>and</strong> pass <strong>and</strong> the <strong>solar</strong> spectral
shape. The error in the measurements of the absolute DU scale can be
estimated as ranging from 3 % to 5 %.
Figure.5.

Depth (DU)
Ozone depth (DU)
340
330
320
310
300
290
280
270
260
250
Jan Apr Months
Aug Dec
1982.
Figure 5: Secular trend by least square tech. for the year 1982
306
304
302
300
298
296
294
292
290
288
1970 1975 1980 1985 1990 1994
Time in year
Figure 6: Secular trend by least square method for April
The secular trends for both the data sets have manifest that the <strong>depletion</strong> at
Pakistan coastal region has got some value albeit the rate is small. With the help
of trend equation we are in the position to predict future status of <strong>ozone</strong> <strong>layer</strong>
<strong>depletion</strong> not only for Pakistan but also for other countries of the region.
Albeit the from satellite can give near-global covering, but can provide only
averaged values of the measured quantity over large regions both in vertical It is
rather difficult as well as expensive to properly h<strong>and</strong>le the working of satellite

instruments because of the variations in the speed <strong>and</strong> orbit of the satellites.
Also, in case of failure of any of the components, necessary repairs take time
that influences the accuracy of data. Whereas, the ground based radars can
provide data with high vertical resolution by measuring small differences in the
time delays of the return pulses above the radar sites.
REFERENCES
(1) Andrews G D (2000), “An Introduction to Atmospheric Physics”, Cambridge
University Press, London
(2) Johnston, H S (1971), “Reduction of stratospheric <strong>ozone</strong> by oxide catalyst
supersonic transport”, Science, 173: 517-522
(3) Mészàros E (1995), “Global <strong>and</strong> Regional Changes in Atmospheric
Composition”, CRC Press Inc., 2000 Corporate Blvd., Florida
(4) Box G E P <strong>and</strong> Jenkins G M (1976),” Time series analysis, forecasting <strong>and</strong>
control”, rev. ed. Holden-Day, San Francisco
(5) Bojkov R, Bishop L, Hill W J, <strong>and</strong> Tiao G C (1990), “A statistical trend
analysis of revised Dobson total <strong>ozone</strong> data over the northern hemisphere”,
J. Geophs. Res. 95: 9785-9808
(6) Stolarski R S, Bloomfield P, Herman J R, (1991), “Total <strong>ozone</strong> trends
deduced from Nimbus 7 TOMS data”, Geophys. Res. Lett. 18: 1015-1018
[7]. Gupta S (1999), “Research Methodology <strong>and</strong> Statistical Techniques”, Deep
& Deep Publications, New Delhi.
[8] Granger C W J, <strong>and</strong> Newbold P (1986), Forecasting Economic Time Series,
2nd.Edn., Academic Press, New York
[9] Priestley M B(1988), Non-linear <strong>and</strong> Non-stationary Time series Analysis,
Academic Press, London
[10] Newbold P (1988), Some recent developments in Time Series analysis,
I,II,<strong>and</strong> III. <strong>Inter</strong>. Statist. Rev., 56, 17-29
[11 ] M.Ayub Khan Yousuf Zai & Jawaid Quamar (1998), Sing. J. Phys., Vol. 14,
# 1, pp 69-79
(12) M.Ayub Khan Yousuf Zai & Jawaid Quamar (2001), “ Study the
phenomenon of <strong>ozone</strong> <strong>layer</strong> <strong>depletion</strong> as a physical process”, Indian. J.
Phys., 75B (4), 307-314.

[13] M.Ayub Khan Yousuf Zai, (2004),”A Quantitative Study of effects of Ozone
Layer Depletion on Marine Organisms with reference to coastal regions of
Pakistan,” A Ph.D. thesis.
[14] Anderson T W (1971), The Statistical Analysis of Time Series, Wiley, New
York
[15] Haggan V <strong>and</strong> Ozaki T (1979), Amplitude-dependent AR model fitting for
non-linear r<strong>and</strong>om vibrations, Paper presentation at the <strong>Inter</strong>national Time
Series meeting, University of Nottingham, UK, March 1979.
[16] Clevel<strong>and</strong> W S <strong>and</strong> McGill (1987), Graphical perception: the visual decoding
of quantitative information on graphical displays of data, J. Roy. Stat. Soc.,
A 150, 300
[17] Basu S K, Sinha B K, & Mukherjee S P (2000), Frontiers in Probability <strong>and</strong>
Statistics, Narosa Publishing House, New Delhi
[18] Subba Rao T (1998), “ Applications of Time Series Ananlysis in Astronomy
<strong>and</strong> Meteorology”, Chapman & Hall, London
[19] Rowl<strong>and</strong> F S (1991), Ann. Rev. Phys. Chem. 42, 731
[20] Bojkov R D <strong>and</strong> Fioletov (1995), Geophys., Res. Lett., 22, 2729
[21] Henriksen K <strong>and</strong> Roldugin V (1995), Geophys., Res., Lett.. 22, 3219
[22] Henriksen K And Larsen H H S (1992), J.Atmos. Terr. Phys., 55, 1113
[23] Herman J R <strong>and</strong> Larko D (1993), J. Geophys. Res., 98, 12783
[24] Hofmann D J, Oltmans J S, <strong>and</strong> Harris J M (1994), Geophys. Res. Lett., 21,
1779
[25] Hofmann D J, Oltmans J S, <strong>and</strong> Koenig G L (1996), Geophys. Res.Lett., 23,
1533
[26] McPeter R D <strong>and</strong> Hll<strong>and</strong>sworth (1996), Geophys. Res. Lett., 23, 3699
[27] Rol<strong>and</strong>o R G (1994), J. Geophys. Res. Vol. 99, # D6. 12,937-12,951
[28] Portmann R W, Solomon S, <strong>and</strong> Rol<strong>and</strong>o R G (1996), J. Geophys. Res.
Vol. 101, # D17, 22,991-23,000
[29] Rol<strong>and</strong>o R G, Frode S, <strong>and</strong> Jeffery T (1992), J. Geophys. Res. Vol. 97, #
D12, 12967-12,991
[30] Rol<strong>and</strong>o R G (1994), Physics World, April, 1994, 49-55
[31] Tatsuo Shimazaki (1989), Geophysics <strong>and</strong> Astrophysics Monograph,
ed., Terra Scientific. Tokyo

[32] Attilio B <strong>and</strong> Sharon B (1997), Encychlopedia of Energy <strong>and</strong> Environmental,
p 198-1225, Wiley, New York
[33] Solomon S, <strong>and</strong> Daniel L A (1992), Nature, 357, 33-37
[34] Prather M J, <strong>and</strong> Watson R T (1990), Nature, 344, 729-734
[35] Kallenrode B M (2004), “ Space Physics”, Springer Verlag, Berlin, New York
[36] Zeilik M (1994), Astronomy, the evolving universe, John Wiley, New York
[37] Richard S L (1990), Dynamics in Atmospheric Physics, Cambridge
University Press, New York
[38] Kaler J B (1994), Astronomy, HarperCollins College Publishers, New York
[39] Zirin H (1989), Astrophysics of the Sun, Cambridge University Press,
New York
[40] Unno W & Yuasa M (1992), Astrophys.<strong>and</strong> Space Sci., 189, 271-278
[41] Seinfeld H J & P<strong>and</strong>is N S (1998), Atmospheric Chemistry <strong>and</strong> physics,
from air pollution to climate, JohnWiley, New York
[42] R<strong>and</strong>el W J & Gille J C (1993), Nature, 365, 533-535

SPACE RADIATION EFFECTS ON SPACECRAFT
MICROELECTRONICS
Dr. Shazia Maqbool <strong>and</strong> Ayesha Javed
Satellite Research <strong>and</strong> Development Center (SRDC) Lahore
SUPARCO, Pakistan
ABSTRACT
Srdclhr@ nexlinx.net.pk
One of the primary design considerations for a space mission is the
survivability of its electronic systems in an ionizing radiation environment. The
increasingly complex mission requirements of modern spacecraft for military,
commercial <strong>and</strong> scientific applications have m<strong>and</strong>ated the need for highly
capable microelectronics. This trend has been accompanied by the need to
provide affordable, lightweight, low volume <strong>and</strong> low power systems. Both these
trends have resulted in the need to adopt commercial microelectronics for
space missions as well as to examine emerging technologies. By their very
nature, commercial technologies are constantly changing.
The effects of radiations on commercial device technologies have widely been
discussed in literature. This paper summarizes these discussions to outline
trends in different radiation effects with changing device technologies. A review
on radiation response of very large scale integration (VLSI) devices will also be
presented.
I. INTRODUCTION
In the past, rad-hard technology derived from military programmes has
dominated the space industry <strong>and</strong> military applications have had significant
influence on the direction <strong>and</strong> capability of the electronic industry. However,
continued reductions in military budgets, coupled with dramatic growth of the
consumer electronics industry, have reduced this influence. Market drivers for
consumer electronics have taken their products <strong>and</strong> supporting technologies
far beyond the capability of specialized military technology. Consequently, radhard
technology lags behind the commercial development by about two
generations. This situation has led space industry to commercial
microelectronics. However, the use of commercial off-the-shelf (COTS)
technology brings new issues on the reliability in the space environment.
One of the most significant trends in commercial microelectronics is the move
towards smaller dimensions – i.e. “scaling”. Scaling has some benefits for
radiation effects resilience: As gate oxide thick-nesses decrease, hole trapping
<strong>and</strong> interface trap build-up are becoming less significant, leading to a trend
toward improved total-ionising dose (TID) performance [13, 4]. Similarly, the
increasing use of buried epitaxial substrates over the last 20 years [14, 30] has
helped decrease single-event latch-up (SEL) sensitivity – both by limiting the
charge-collection volume, <strong>and</strong> by decreasing the substrate series resistance

[35, 29, 3]. In addition, the concomitant trend towards reduced supply voltage
levels suggests that device threshold voltages should soon fall below that
required sustaining latch-up. Indeed, if continued scaling forces the industry to
adopt silicon-on-insulator (SOI) technology, latch-up should soon be eliminated
[13, 30].
However, whilst some radiation effects are becoming less of an issue, new
threats have emerged: Another significant trend in COTS technology is the
move towards the use of “smart” logic, in device architectures, e.g. advanced
memories, complex micro-controllers <strong>and</strong> field-programmable gate arrays
(FPGAs), etc. Here the performance of the logic device requires the inclusion
of complex control circuitry internal to the die. Whilst transparent to the user,
such circuitry may offer a significant target to ionizing particles <strong>and</strong> thus be
prone to single-event effects (SEEs), such as a single-event upset (SEU) or
single-event transient (SET). In such circuitry, these can manifest themselves
as single event functional interrupts (SEFIs), whereby the device exhibits an
unexpected change in its observable output state [18].
II. SINGLE EVENT EFFECTS (SEE)
SEEs are the result of free-charge generation in semiconductor materials
caused by a single ionizing particle strike. Different phenomena may arise
depending on the location <strong>and</strong> the instant of the particle strike.
II.I. Single Event Upset (SEU)
Single-Event Upset (SEU) is a non-destructive single-event-effect. These are
soft errors associated with changes in the state of memory elements – i.e.
unexpected bit-flips due to particle strikes. Re-writing the data will restore the
memory states.
As technologies moved towards smaller feature sizes <strong>and</strong> thus had less charge
stored on their circuit nodes, it was expected that the critical charge (<strong>and</strong> thus
the linear energy transfer (LET) threshold) for SEU would decrease [1]. Indeed
this trend was observed in the 1980’s in the now older generation of devices.
However, results on more recent technologies, with minimum feature sizes
ranging from 3μm to 0.35μm do not follow this trend; in fact, the threshold LET
has remains approximately constant. This levelling off at a minimum LET has
been postulated to be the result of manufacturers taking into account the need
to maintain a low upset rate from naturally-occurring radio<strong>activity</strong> (e.g. alpha
particles) or atmospheric neutrons during the design <strong>and</strong> manufacture of their
parts [13, 4]. Recent SEE testing of complementary metal oxide semiconductor
(CMOS) technologies (0.15 μm, 0.13 μm, <strong>and</strong> 90 nm) conforms with these
predictions <strong>and</strong> it is found that the decreased size of the shrinking latches
combined with robust design efforts has stabilized the critical neutron cross
sections of state of the art static latches. Indeed, the mean time between failure
(MTBF) calculated for a million logic gate FPGA fabricated with a 90-nm
technology is better than that of the MTBF of a similar complexity array
fabricated in 150- or 130-nm processes.

Figure 1: Relationship between Feature Size <strong>and</strong> Critical Charge, [31]
II.II. Single Event Transients (SET)
This is are also a class of non-destructive soft-error that can cause changes of
logical state in combinational logic, or may be propagated in sequential logic
through “glitches” on clock or set/ reset lines, etc.
So far, it has been observed that radiation response of logic devices is
dominated by errors in registers <strong>and</strong> memory cells – i.e. SEUs [14]. However,
as devices are further scaled down to smaller feature sizes <strong>and</strong> faster speeds,
soft errors in combinational logic, SETs, are expected to become more
probable. In contrast to SEUs, which do not show frequency dependence,
SETs depend significantly on the operating speed of the devices in question [4,
14]. According to models developed by Shivakumar et al., soft errors in
combinational logic are likely to become comparable to SEUs in unprotected
memory elements by the year 2011 [38].
II.III. Single Event Latch-Up (SEL)
Single Event Latch-Up (SEL) is a potentially destructive single event effect. It
can occur due to the presence of parasitic bipolar transistors inherent in CMOS
structures, <strong>and</strong> may lead to an effective short circuit through the device. There
are two types of SEL: In traditional or destructive SEL, the device current
consumption increases immediately above the maximum specified for the
device, causing permanent damage to the device through the burn-out of bondwires,
etc. The other type is known as a micro-latch-up, <strong>and</strong> during such an
event, the device current increases above the normal operating voltage but not
above the maximum specified. Micro-latch-up halts device operation <strong>and</strong> a
power reset is required to recover the device. The device may still operate but
show signs of permanent damage, such as exhibiting a step-change in its
current consumption. Alternatively the damage may not be immediately
apparent. Sometimes, a series of micro-latch-up events in quick succession
can lead to destructive SEL.
II.IV. Single Event Functional <strong>Inter</strong>rupt (SEFI)
One of the most significant trends in COTS technology is the move towards
device architectures with in-built “smart logic” – that is for devices which have
internal operations transparent to the user. Smart logic requires the inclusion of
more control circuitry internal to the die. For example, flash memory devices

contain a state machine that generates the signals needed to perform
read/write operations. Many dynamic r<strong>and</strong>om-access memory (DRAM)
devices, microprocessors <strong>and</strong> field-programmable gate-arrays (FPGAs) have a
built-in self-test capability, which again increases device complexity. Many
DRAMs include redundant memory rows <strong>and</strong> columns intended to increase
device reliability. When the DRAM is powered up, weak or bad rows or
columns are replaced with their redundant counterparts <strong>and</strong> a redundancylatch
has to be used to hold the device configuration. Many synchronous-
DRAM (SDRAM) devices also have internal state machines that implement
pipelines, programmable refresh modes <strong>and</strong> various power states. In most of
these cases, this control logic is not directly accessible to the user.
Experience from space flight <strong>and</strong> ground-based ion-beam testing of many
COTS devices reveals that this internal control circuitry is susceptible to
radiation effects, often manifesting itself as the class of single event effects
(SEEs) called single event functional interrupts (SEFIs), whereby an internal
upset in this logic causes an unexpected change in the observable state of the
device. SEFI was first mentioned in 1996 [5, 18], although SEFI-like signatures
had been reported earlier [17, 42, 19]. Early reports were confined to
microprocessor SEFIs, however, emerging data h<strong>and</strong>ling devices, such as
advanced memories <strong>and</strong> FPGAs, have also been found to be susceptible.
III. MULTIPLE BIT UPSET (MBU)
Sometimes, more than one memory cells can be corrupted by the charge
deposition originating from the same particle. Such events are characterized as
single-event multiple-bit upsets. These errors are mainly associated with
memory devices, although any register is a potential target. MBUs, which have
been observed during ground-based ion-beam testing <strong>and</strong> in during space
flight, may be attributed to three mechanisms:
• A single particle strike affects more than one bit in one or more words;
• Coincidental, independent SEUs (from different particle strikes) appear
as an MBU;
• A single SEFI event causes MBUs (termed as block SEFIs).
The second mechanism is not in fact a true MBU, but is instead an artifact of
the diagnostic algorithm <strong>and</strong> the underlying (usually high) SEU rate. Both
SRAMs <strong>and</strong> DRAMs are prone to MBUs, with DRAMs, generally having a
higher probability of getting such errors. MBU rates are expected to increase
with the device scaling. However, in some cases, less dense devices have
been found to show higher MBU rates than more dense devices [39]. On the
other h<strong>and</strong>, reports of block SEFI events are increasing. Floating gate
memories, which do not exhibit traditional MBUs, are susceptible to this class
of MBUs.
IV. TOTAL IONISING DOSE
Total ionizing dose in semiconductor devices depend on the creation of
electron-hole pairs within dielectric <strong>layer</strong>s <strong>and</strong> subsequent generation of traps
at or near the interface with the semiconductor. This can produce a variety of
device effects such as flat-b<strong>and</strong> <strong>and</strong> threshold voltage shifts, surface leakage
currents, <strong>and</strong> speed degradation of the devices. Many current COTS parts are
very susceptible to total dose damage, <strong>and</strong> may fail at total-dose levels of 5
krad (Si) or even less. Total dose is therefore an issue in space flight where

long mission lifetimes, <strong>and</strong>/or exposure to the Van Allen radiation belts means
that these dose levels can be easily exceeded.
As commercial devices are scaled to smaller feature sizes, so gate oxide
thicknesses are also decreasing. The threshold voltage shift that results from
hole traps in gate oxides decreases as the square of oxide thickness,
assuming that the hole trapping efficiency is unchanged as the oxide thickness
is reduced [23]. It has been shown that hole trapping will reduce further
because of tunneling as oxides are scaled below 10nm [34]. Thus, for the
highly scaled devices, hole trapping <strong>and</strong> interface traps are not likely to remain
major concerns for devices operating in a radiation environment. However, it
should be noted that even in highly scaled devices there would be isolation
dielectrics, which are still relatively thick. Thus this is isolation/field oxide that
dominates total dose response of commercial devices [39, 13, 4,].
V. TRENDS IN DEVICES
This section presents a review of radiation issues in SOTA <strong>and</strong> emerging
microelectronics.
V.I. MEMORIES
Space radiation effects on different types of memory devices are discussed
below.
V.I.I. DRAMs
DRAM technology has grown rapidly in recent years. The first DRAM,
introduced in the early 1970s, contained only 1,024 bits, but modern DRAMs
are available containing 256 megabits or more. Large DRAMs incorporate a
controller to provide parallelism in operations <strong>and</strong> hence an increased memory
throughput. Their high density <strong>and</strong> superior performance characteristics make
them an attractive c<strong>and</strong>idate for on-board data recording to fulfill increasingly
higher mission data storage requirements. In addition, recent technology has
permitted the stacking of dies into single units providing even denser
packaging of DRAMs into a single device. However, the use of DRAMs for
space applications does not come without certain penalties. DRAMs have
historically been considered as devices very sensitive to SEUs, since the circuit
errors were observed on these components back in the seventies. Their high
sensitivity to ionising particle strikes comes from their characteristic of
passively storing bit information as charge stored in a circuit node. Although
the amount of charge stored decreases steadily from one technology
generation to the next, unexpectedly, the error rate has been found to
decrease or stay constant [24, 7, 14]. However, the observation of nontraditional
SEEs, e.g. SEFIs, raises new concerns over the reliability of these
devices in radiation environment. Both the Cassini <strong>and</strong> Hubble Space
Telescope (HST) missions have used solid state data recorders (SSDRs)
made from DRAM [21]. Observed SEE results include the st<strong>and</strong>ard single bit
upset (SEU) errors, stuck bits, MBUs, <strong>and</strong> column or row errors (where a single
ion strike induces a partial or full address column or row to be in error in a
single device) [21, 39]. Label termed column <strong>and</strong> row errors as block SEFIs
[20]. Ground-based ion-beam testing of DRAMs has also shown st<strong>and</strong>ard

SEFIs (e.g with the device entering test or st<strong>and</strong>by modes) in addition to
above-mentioned errors [20, 18].
It should be noted that advanced DRAM architectures are also emerging. An
example of an advanced DRAM is Synchronous DRAM (SDRAM) that can
provide even more storage capacity on a chip along with an increase in access
speed. These devices have underlying characteristics of stuck bits <strong>and</strong> MBUs.
Increased level of complexity in these devices leads to higher susceptibility of
SEFIs [33].
V.I.II. SRAMs
These devices are susceptible to SEUs, generally with quite low thresholds
(often below 1 MeVcm2/mg) [2]. They are also subject to MBUs. However, their
MBU rates are significantly smaller than their SEU rates. There is no evidence
from in-orbit observations that the denser devices are necessarily more prone
to MBUs than the older devices [39, 41] although ground based tests suggest
this. Stuck bits are sometimes observed, but only during irradiation with
particles of high LET (i.e. heavy-ions) [7]. Newer generations of SRAMs, using
a 6T cell design, are expected to have an improved SEU <strong>and</strong> total dose
response [22, 32, 7].
V.I.III. Flash Memories
Flash memory technology comes with advantages in terms of density <strong>and</strong> nonvolatility.
It provides fast read access, comparable to that of DRAMs. However,
erasing <strong>and</strong> writing new data in flash memories is a more complex operation,
requiring application of high voltage. Newer flash devices use complex internal
architectures to improve write <strong>and</strong> erase times, as well as to make device
operation more transparent to user.
There are two types of flash architecture: NOR <strong>and</strong> NAND. Whilst both
architectures have the same basic storage element, it is the interconnection of
these memory cells that distinguishes structure. In the NOR flash memory
array, the bit line logic goes to “0” if any of the memory cell transistors is on. In
the NAND flash memory array, all memory cell transistors are required to be on
for bit line to go to “0” state. The NAND architecture is more sensitive to
radiation [25].
Radiation testing has been performed on Intel, Samsung, Toshiba, AeroFlex,
<strong>and</strong> SanDisk Flash memory devices [36, 28, 27]. The observed SEEs are
dominated by errors in their complex internal architectures rather than in the
non-volatile storage elements. Test results have shown that flash devices are
immune to SEUs when powered off. The probability of getting an upset is
particularly high during write mode. Static <strong>and</strong> dynamic read modes have also
demonstrated complex control-related upsets, where the functional
consequences of these errors can be multiple. Sometimes, a steep increase in
the current consumption is observed during or after irradiation.
V.I.IV. EEPROMs
Electrically Erasable Programmable Read-Only Memory (EEPROM) provides
non-volatile storage like Flash memories. The principal difference between the
two technologies is that an EEPROM requires data to be written or erased one

yte at a time whereas flash memory allows data to be written or erased in
blocks. This makes flash memory faster. Radiation test results on EEPROMs
are presented in [18, 20]. Similar to flash memories, the devices are found to
be more susceptible during write mode. SEFI has been observed in these
memories as well.
V.I.V. Mitigations for Memory Devices
There are three approaches to improve survivability of microelectronics in
space environment.
• Firstly, to invent better semiconductor process to make the memory
cell less susceptible to direct radiation. Researchers found that performance
under radiation can be improved by repositioning various component junctions
or by using silicon-on-sapphire (SOS) or SOI technologies. All these methods
require redesign of the memory chips <strong>and</strong> call for new process technologies
that are not widely used for memory chips <strong>and</strong> therefore, present a risk in
production cost <strong>and</strong> in scalability when memory technology changes.
• The other way is to use COTS memory parts. It has been found that
every memory chip exhibits slightly different characteristic under radiation
environment. Even the same batch of chips can react quite differently among
each other. Some chip works well in the energy environment while others fail
miserably. Therefore, the most suitable way is to test-<strong>and</strong>-select. NASA
decided to use COTS parts even in the <strong>Inter</strong>national Space Station. This is
obviously for cost reason <strong>and</strong> to allow upgrade path as memory technology
advances during their course [43].
• Implementation of EDAC (Error Detection <strong>and</strong> Correction Codes)
techniques in both hardware <strong>and</strong> software can prove helpful in fighting against
the hazardous effects of radiations.
V.II. MICROPROCESSORS
The most complex type of digital logic device used in on-board is the
microprocessor. Using commercial processors for space missions brings
potential advantages in terms of cost <strong>and</strong> design time reduction. These devices
are accompanied by COTS software applications, development tools, <strong>and</strong>
operating systems. Processors such as the 80Cx86 family <strong>and</strong> the Power PC
are being flown in space despite being susceptible to ionizing radiation. These
processors are not radiation hardened <strong>and</strong> have been found to be susceptible
to SEUs.
Microprocessors work by continuously moving data between registers; the
memory unit with arithmetic <strong>and</strong> logic unit operations <strong>and</strong> flag registers updates
etc. It is unlikely that all sections will be in use during the processing of a
program. The application software, which is being executed, determines how
many sections are being used at any given time. Moreover, each section might
have a different sensitivity to SEU. Therefore, the SEU-induced effects in
microprocessors are strongly application dependent. Many commercial
processors have been tested to evaluate their suitability for space applications.
These include the 80x86 architectures, SPARC, Motorla 68020, Pentium,
Alpha, <strong>and</strong> PowerPC’s [6, 26, 20, 11, 37, 12]. All devices experienced some
type of SEFI or device lockups at fairly low threshold LETs <strong>and</strong> authors have
suggested that reliable operation in the radiation environment will require at
least a detection <strong>and</strong> reset capability.

One result, which was certainly unexpected, is that the threshold LET has
essentially not changed during last many years [13]. Table1 compares
threshold LET of different microprocessors.
Mitigation:
Implementation of either offline or online fault detection techniques like
watchdog timers, Locksteps <strong>and</strong> built in testing in microprocessors can
help against the radiation caused damages.
Device Manufacturer Feature Size
(approx)
Threshold LET
(MeV-cm 2 /mg)
Z-80 Zilog 3μm 1.5-2.5
8086 Intel 1.5μm 1.5-2.5
80386 Intel 0.8μm 2-3
68020 Motorola 0.8μm 1.5-2.5
LS64811 LSI 1.2μm 2-2.5
90C601 MHS 1.2μm 2-2.5
80386 Intel 0.6μm 2-3
PC603e Motorola 0.4μm 1.7-3
Pentium Intel 0.35μm 2-3
Power PC750 Motorola 0.25μm 2-2.5
PowerPC7455 Motorola 0.18μm 1
Table1: Upset Threshold of Microprocessors [13]
V.III. FIELD PROGRAMMABLE GATE ARRAYS (FPGAs)
This technology offers number of advantages including a highly compact
solution, high integrity, flexibility, reduced cost, faster <strong>and</strong> cheaper prototyping,
<strong>and</strong> reduced lead-time before flight. The capacity <strong>and</strong> performance of FPGAs
suitable for space flight have been increasing steadily for more than a decade.
The application of FPGAs has moved from simple glue logic to complete
platforms that combine several real-time system functions on a single chip [9].
The choice of type of storage for the configuration drives the radiation
performance of the device [15]. FPGAs can be split into two categories: namely
re-programmable <strong>and</strong> one time programmable (OTP). Reprogrammable
technology offers volatile SRAM or non-volatile EEPROM/Flash cells to hold
the device configuration. SRAM-based technology has been evolving at a
faster pace than OTP technology, <strong>and</strong> now features a million system gates or
more on a single chip, <strong>and</strong> hence has become an attractive choice for high
performance applications. However, one must be aware not only of radiation

issues, but also of concerns such as the total loss of configuration from single
ion hit. OTP technology on the other h<strong>and</strong> offers radiation performance
comparable to application specific integrated circuit (ASICs). The most
commonly used OTP FPGAs use antifuses for storing its configuration, either
using oxide-nitride-oxide (ONO) or metal-to-metal (M2M) antifuse structures
[16].
FPGAs are available at different reliability levels, ranging from inexpensive
COTS devices to high reliability, rad-hard devices, as well as different device
speeds, capabilities, <strong>and</strong> configurations. Both Actel <strong>and</strong> Xilinx have been
offering radiation tolerant devices for space applications. The Actel OTP
devices have been the primary FPGA technology used in space flight systems
to date. Actel’s ACT1, ACT2 <strong>and</strong> ACT3 have been very popular among space
designers [15]. Commercial SX <strong>and</strong> SX-A series have also been reported as
being capable of tolerating total doses as high as 100 krads(Si) or more. These
devices are latch-up immune as well. The flip-flop element of the basic device
is considered as being radiation “tolerant”; enhanced radiation hardened levels
can be achieved via incorporation of software or hardware implemented
mitigation strategies [16]. Although ONO antifuses are susceptible to SEGR,
the threshold is found to be quite high [15]. The probability of an SEGR event is
low <strong>and</strong> one rupture does not cause permanent damage to the device.
Typically it would take at least 10 ruptures to cause a failure. Actel introduced
products using a M2M antifuse. It is reported that this antifuse technology is
immune to SEGR effect to a currently tested LET threshold of 80 MeV-cm 2 /mg.
Actel Corporation offers two classes of rad-hard FPGAs, RH1020 <strong>and</strong> RH1280
with equivalent gate densities of 2,000 <strong>and</strong> 8,000 respectively. These products
are latchup immune <strong>and</strong> can withst<strong>and</strong> total dose in excess of 300 krads (Si).
Its SEU performance is predicted to be 10-6 upsets per bit-day in a 90% worstcase
geosynchronous earth orbit. Actel proposes mitigation techniques to
improve SEU performance of these devices by incorporating TMR or by
avoiding flip-flops in the sequential units.Adaptation of one of these techniques
is recommended for critical design sections.
Another rad-hard FPGA RHAX250-S device technology offers a gate count of
250, 000 with even better radiation performance. In addition to these rad-hard
products, Actel offers radiation tolerant devices such as RTAX_S. This device
technology has TID response comparable to RH1020 <strong>and</strong> RH1280 devices,
<strong>and</strong> latchup <strong>and</strong> SEU immunity to 104 <strong>and</strong> 60 MeV-cm2/mg respectively.
Xilinx offers the XQR series for space applications. This series is a subset of
their commercial XC40000XL family. Although these devices have
demonstrated acceptable total dose <strong>and</strong> latch-up performance for many space
applications, they are susceptible to single event effects including SEFIs [8, 9].
Work is in progress at Xilinx for the development of the single event immune
reconfigurable FPGA (SIRF).
SEUs in reprogrammable FPGAs can be grouped into three categories [9]:
Configuration upsets are defined as the upsets in the configuration memory of
the device <strong>and</strong> can be detected by read-back. The likelihood of failure will

depend upon the upset location <strong>and</strong> the specific design utilisation of the device
resources. A configuration upset can sometimes lead to high current states.
For example, a SEU may cause two output drivers to be connected together.
User logic upsets represent upsets in the circuit elements that are not directly
accessible by read-back. The effect of these is dependent upon the particular
logical design implemented by the user.
System upsets are those upsets that occur in the control circuitry of the FPGA.
As an example, it is possible for a single upset in the configuration control
circuitry to change many configuration bits simultaneously .This kind of
anomaly can only be detected by noting the malfunction of the device.
V.IV. GaAs DEVICES
GaAs devices are used in high speed digital <strong>and</strong> microwave applications. Since
no dielectric <strong>layer</strong>s are used in GaAs devices, they exhibit extreme radiation
tolerance - up to several hundred Mrad (GaAs) (two orders of magnitude higher
than for equivalent Si-based technologies) [44].
In general GaAs technology is very susceptible to SEU but hardening with low
temperature buffer <strong>layer</strong>s offers orders of magnitude improvement.
V.V. OPTO-ELECTRONIC DEVICES
Broadly opto-electronic devices include light emitting diodes (LEDs), laser
diodes, optical detectors, <strong>and</strong> optical couplers. These products are also not
considered radiation hard without special processing. For example opto
couplers can be sensitive to degradation of current transfer ratio (CTR) at a
dose of 2-3 krad (Si). Some types of opto-emitters are severely degraded by
the proton bombardment at a TID of 1-2 krads. Displacement damage dose
(DDD) is most pronounced in optocouplers <strong>and</strong> lasers. High b<strong>and</strong>width
optocouplers are extremely sensitive to SEUs. Permanent damage can be
caused by displacement damage, for example loss in LED power (though
devices with a hardened LED will ultimately suffer from photo-response
degradation) [45] <strong>and</strong> from ionizing damage in the photo detector.
Optocouplers can be of linear or digital type <strong>and</strong> many different designs exist,
each with a different radiation response. Optocoupler failures occurred on the
Topex-Poseidon mission [45] at TID levels of 20-30 krad (Si). For systems that
use lens coupling, radiation darkening of the lens, due to TID, can sometimes
be a problem (particularly for graded index, GRIN, lenses).
VI. CONCLUSIONS
Adoption of COTS devices <strong>and</strong> st<strong>and</strong>ards has become a prevailing practice for
space applications. Continuously changing COTS technologies impose a threat
to their reliable use in space radiation environment.
A general trend towards an improved TID <strong>and</strong> SEL response has been
observed. SEFIs <strong>and</strong> SETs are recognized as two potentially dominating
radiation hazards from current knowledge of device technology <strong>and</strong> its
radiation performance. SETs are predicted to be of serious concern in near

future. SEFIs, on the other h<strong>and</strong>, have been reported with present technology
<strong>and</strong> likely to get worse as device technology moves towards smarter <strong>and</strong> more
transparent architectures.
Over the past many years, COTS technology has been flown in space systems
with a reasonable expectation of success. This has primarily been achieved by
a thorough underst<strong>and</strong>ing of radiation response of these devices <strong>and</strong> then,
application of appropriate mitigation techniques <strong>and</strong> sound engineering
practices.
VII. ACKNOWLEDGMENT
This work is attributed to the unmatched guidance of Dr. Riaz M Suddle <strong>and</strong> to
SUPARCO for sponsoring it.
VIII. REFERENCES
[1] Akers, L.D., “Microprocessor Technology <strong>and</strong> Single Event
Suscptibility”, Proc. 10th Annual AIAA/USU Conference on Small
Satellites, Logan, Utah, 1996
[2] J.R.Coss et al., “Device SEE Susceptibility Update: 1996-1998”, IEEE
NSREC Data Workshop Record, pp. 60-81, July, 1999
[3] P.E. Dodd, M.R Shaneyfelt, E. Fuller, J.C Pickel, F.W. Sextan <strong>and</strong> P.S.
Winokur “Impact of Substrate thickness on Single-Event Effects in
Integrated Circuits”, IEEE Trans. Nucl. Sci., Volume-48, Issue-6, pp.
1865-1871, December, 2001
[4] Paul V. Dressendorfer “Basic Mechanisms for the New Millennium”,
IEEE NSREC Short Course, July, 1998
[5] EIA/JEDEC St<strong>and</strong>ard, “Test Procedures for the Measurement of Single
Event Effects in Semiconductor Devices from Heavy Ion Irradiation”,
Electronic Industries Association, Engineering Department, Arlington,
1996
[6] Estreme, F. et al., “SEU <strong>and</strong> Latchup Results for SPARC Processors”,
IEEE Radiation Effects Data Workshop Record, pp. 13-19, July, 1993
[7] Faccio, F., “COTS for the LHC Radiation Environment: The Rules of
the game”, 6th Workshop on Electronics for LHC Experiments,
Krakow, Pol<strong>and</strong>, 2000
[8] Fuller, E. et al., “Radiation Characterization, <strong>and</strong> SEU Mitigation, of the
Virtex FPGA for Space-Based Reconfigurable Computing”, Presented
at IEEE NSREC Conference, October, 2000
http://klabs.org/richcontent/Papers/Xilinx/Xilinx_NSREC2000.pdf
accessed on 5th March 2004
[9] Gaisler, J., “Suitability of Reprogrammable FPGAs in Space
Applications”, ESA Technical Report, 2002
http://www.estec.esa.nl/microelectronics/asic/fpga_002_01-0-4.pdf
accessed on 5th March 2004
[10] Hodgart, M.S. <strong>and</strong> Tiggeler, H., “Introduction to error Correcting Codes
in Satellite Applications”, Technical Report, Surrey Space Centre, 1999
[11] Howard, J.W. et al., “Total Dose <strong>and</strong> Single Event Effects Testing of
the Intel Pentium III (P3) <strong>and</strong> AMD K7 Microprocessors”, IEEE
Radiation Effects Data Workshop Record, pp. 38-47, July, 2001

[12] Irom, F. et al., “Single-Event Upset in Commercial Silicon-On-Insulator
PowerPC Microprocessors”, IEEE Trans. on Nucl. Sci., Vol. 49, Issue-
6, pp. 3148-3155, December, 2002
[13] A.H. Johnston, “Radiation Effects in Advanced Microelectronis
Technologies”, IEEE Trans. Nucl. Sci., Volume-45 , Issue-3 , pp. 1339-
1354, June, 1998
[14] Johnston, A.H., “Scaling <strong>and</strong> Technology Issues for Soft Error Rates”,
4th Annual Research Conference on Reliability, Stanford University,
October, 2000
http://nepp.nasa.gov/docuploads/40D7D6C9-D5AA-40FC-
829DC2F6A71B02E9/Scal-00.pdf accessed on 5th March 2004
[15] Katz, et al., “Radiation effects on Current Field Programmable
Technologies”, Presented at IEEE NSREC Conference, July, 1997
http://klabs.org/richcontent/Papers/nsrec97fpgapaper.PDF accessed
on 5th March 2004
[16] Katz, R. et al., “Logic design Pathology <strong>and</strong> Space Flight Electronics”,
MAPLD <strong>Inter</strong>national Conference Proceedings, 1999
http://klabs.org/richcontent/Papers/Mapld99/A4_Katz_P.pdf accessed
on 5th March 2004
[17] Koga, R. et al., “Techniques of Microprocessor Testing <strong>and</strong> SEU rate
Prediction”, IEEE Trans. on Nucl. Sci., Vol.-32, Issue- 6, pp. 4219-
4224, December, 1985
[18] Koga, R. et al., “Single Event Functional <strong>Inter</strong>rupt (SEFI) Sensitivity in
Microcircuits”, Proc. 4th European Conference on Radiation <strong>and</strong> its
Effects on Components <strong>and</strong> Systems (RADECS), pp. 311-318, 1997
[19] Label, K. et al., “SEU Tests of 80386 Based Flight-Computer/Data
H<strong>and</strong>ling System <strong>and</strong> of Discrete PROM <strong>and</strong> EEPROM Devices <strong>and</strong>
SEL tests of Discrete 80386, 80387,PROM, EEPROM, <strong>and</strong> ASICs”,
IEEE Radiation Effects Data Workshop Record, pp. 1-11, 1992
[20] Label, K. et al., “Radiation Effect Characterization<strong>and</strong> Test Methods of
Single Chip <strong>and</strong> Multi Chip Stacked 16Mbit DRAMs”, IEEE Trans. on
Nucl. Sci. Vol 43, Issue-6, pp. 2974- 2981, December, 1996
[21] Label, K., “Single Event Effect Test Results for C<strong>and</strong>idate Spacecraft
Eectronics”, IEEE Radiation Effects Data Workshop, pp. 14-21, July,
1997
[21] Label, K. et al., “Anatomy of an In-Flight Anomaly: Investigation of
Proton-Induced SEE Test Results for the Stacked IBM DRAMs”, IEEE
Trans. on Nucl. Sci., Vol 45, Issue-6, pp. 2898-2903, December, 1998
[22] A.J.Lelis et al., “Radiation Response of Advanced Commercial
SRAMs”, IEEE Trans. Nucl. Science, Vol.43, Issue-6, pp. 3103-3108,
December, 1996
[23] T.P. Ma <strong>and</strong> P.V. Dressendorfer, “Ionizing radiation effects in MOS
devices <strong>and</strong> circuits”, Wiley <strong>Inter</strong>-science, New York, 1989
[24] Massengill, L.W, “Cosmic <strong>and</strong> Terrestrial Single Event Radiation
Effects in Dynamic R<strong>and</strong>om Access Memories”, IEEE Trans. on Nucl.
Sci, Vol. 42, Issue-2, pp. 576-593, April, 1996
[25] Miyahira, T. <strong>and</strong> Swift, G., “Evaluation of Radiation Effect in Flash
Memories”, JPL Publication, 2003
[26] Moran, A. et al., “Single Event Effect testing of the Intel 80386 Family
<strong>and</strong> the 80486 Microprocessor”, Proc.3rd European Conference on
Radiation <strong>and</strong> its Effects on Components <strong>and</strong> Systems (RADECS), pp.
263-269, September, 1995
[27] Nguyen, D.N <strong>and</strong> Scheick, L.Z., “SEE <strong>and</strong> TID of Emerging Non-
Volatile Memories”, JPL Publication, 2002

[28] Nguyen, D.N. et al., “Radiation Effects on Advanced Flash Memories”,
IEEE Trans. on Nucl. Sci., Vol 46, Issue-6, pp. 1744 – 1750,
December, 1999
[29] A. Ochoa, Jr., <strong>and</strong> P.V. Dressendorfer, “A Discussion of the Role
Distributed Effects in Latchup,” IEEE Trans. Nucl. Sci., vol. NS-28, pp.
4292–4294, Dec. 1981
[30] Oldham, T.R., “How device scaling affects single event effects
sensitivity”, IEEE NSREC Short Course, July, 2003
[31] Petersen, E.L et al., “Calculations of Cosmic Rays Induced Soft Upset
<strong>and</strong> Scaling in VLSI Devices”, IEEE Trans. Nucl. Sci., Vol-29, Issue-6,
pp. 2055-2063, December, 1982
[32] C.Poivey et al., “Radiation Characterisation of Commercially Available
1Mbit/4Mbit SRAMs for Space Applications”, NSREC Data Workshop,
1998
[33] Rodgers, J et al., “Advanced Memories for Space Applications”, IEEE
Microelectronics Reliability & Qualification Workshop, December, 2001
http://parts.jpl.nasa.gov/mrqw/mrqw_presentations/S1_rodgers.pdf
accessed on 5th March 2004
[34] N.S. Saks, M.G. Ancona <strong>and</strong> J.A. Modola, “Radiation Effects in MOS
Capacitors with Very Thin Oxides at 80K”, IEEE Trans. Nucl. Sci., Vol.
31, pp. 1249, 1984
[35] J.E. Schroeder, A. Ochoa, Jr., <strong>and</strong> P.V. Dressendorfer, “Latchup
elimination in bulk CMOS LSI circuits,” IEEE Trans. Nucl. Sci., Vol-27,
pp. 1735–1738, December, 1980
[36] Schwartz, H.R. et al., “Single Event Upset in Flash Memories”, IEEE
Trans. on Nucl.Sci., Vol-44, Issue-6, pp 2315-2324, December, 1997
[37] Seifert, N. et al., “Impact of Scaling on Soft-Error Rates in Commercial
Microprocessors”, IEEE Tans. on Nucl. Sci., Vol. 49, Issue-6, pp. 3100-
3106, December, 2002
[38] Shivakumar, P. et. al, “Modeling the Effect of Technology Trends on
Soft Error Rate of Combinational Logic”, Proc. <strong>Inter</strong>national
Conference on Dependable Systems <strong>and</strong> <strong>Network</strong>s (DSN), pp. 389-
398, June, 2002
[39] Swift, G.M., <strong>and</strong> Guertin, S.M., “In-flight Observations of Multiple Bit
Upset in DRAMs”, IEEE Trans. on Nucl. Sci., Vol. 47, Issue-6, pp.
2386-2391, December, 2001
[40] Underwood, C.I., “Single Event Effects in Commercial Memory Devices
in the Space Radiation Environment”, PhD Thesis, University of
Surrey, 1996
[41] Underwood, C.I., “Observations on the reliability of COTS-devicebased
SSDRs operating in low earth orbit”, Fifth European conference
on Radiation <strong>and</strong> its Effects on Components <strong>and</strong> Systems (RADECS),
pp. 387-393, September, 1999
[42] Velazco, R. et al., “SEU Testing 32-bit Microprocessors”, IEEE
Radiation Effects Data Workshop Record, pp. 16-20, July, 1992
[43] Products guide www.Simmtester.com. Radiation hardened memory
devices
[44] C Claeys <strong>and</strong> E Simoen, Literature study on radiation effects in
advanced semiconductor devices, ESTEC contract report P35284-IM-
RP-0013, March 30 2000
[45] A H Johnston <strong>and</strong> B G Rax, "Proton damage in linear <strong>and</strong> digital
optocouplers," IEEE Trans Nucl Sci, vol 47, pp 675 -681, 2000.
[46] http://klabs.org/richcontent/MAPLDCon98/Papers/c4_miyahira.pdf
accessed on 5th March 2004
[47] http://parts.jpl.nasa.gov/docs/PID16621.pdf
accessed on 5th March,
2004