RESEARCH ON NONLINEAR CONDUCTIVITY- TEMPERATURE ... · material. High power electrons usually...
Transcript of RESEARCH ON NONLINEAR CONDUCTIVITY- TEMPERATURE ... · material. High power electrons usually...
POLITECNICO DI MILANO
Scuola di Ingegneria Industriale e dell'Informazione
Corso di Laurea Magistrale in Ingegneria Elettrica
RESEARCH ON NONLINEAR CONDUCTIVITY-
TEMPERATURE CHARACTERISTIC OF POLYIMIDE
MODIFIED BY ZNO AND ITS TRAP DISTRIBUTION
Relatore: Prof. GIOVANNI DOTELLI
Correlatore: Doc. SAVERIO LATORRATA
Tesi di Laurea Magistrale di:
LI Kangning
Matr. 835974
Anno Accademico 2016-2017
ACKNOWLEDGEMENTS
1
Indice
Indice ........................................................................................................................................... 1
Acknowledgements .................................................................................................................... 4
Indice Delle Figure ..................................................................................................................... 5
Indice Delle Tabelle ................................................................................................................... 6
Abstract ....................................................................................................................................... 7
Abstract (Italiano) ...................................................................................................................... 8
1 Introduction .................................................................................................................. 9
1.1 Introduction of Deep Dielectric Charging ................................................................. 9
1.1.1 What is Deep Dielectric Charging ............................................................... 9
1.1.2 How to Solve Deep Dielectric Charging ................................................... 10
1.1.3 Influence of Grounding Condition ............................................................. 10
1.2 Contents of work........................................................................................................ 12
2 Theory Background ................................................................................................... 13
2.1 Conductivity in Space ............................................................................................... 13
2.2 Classic Conducting Theory ....................................................................................... 14
2.2.1 Hopping Conduction ................................................................................... 14
2.2.2 Schottky Effect ............................................................................................ 14
2.2.3 Poole-Frankel Effect ................................................................................... 15
2.2.4 Tunneling Effect.......................................................................................... 16
2.2.5 Space Charge Limited Current Effect ........................................................ 16
3 Modification of Polyimide ........................................................................................ 19
3.1 Develop of Modification Technology ...................................................................... 19
3.2 Main Features of Raw Materials............................................................................... 22
3.2.1 Polyimide ..................................................................................................... 22
ACKNOWLEDGEMENTS
2
3.2.2 ZnO .............................................................................................................. 23
3.2.3 Coupling Agent ........................................................................................... 24
3.3 Modification Process ................................................................................................. 25
3.3.1 Powder Mixing ............................................................................................ 25
3.3.2 Compression Molding................................................................................. 26
4 Basic Dielectric Performance ................................................................................... 28
4.1 Volumn Resistivity .................................................................................................... 28
4.2 Dielectric Spectrum ................................................................................................... 29
4.2.1 Relative Permittivity ................................................................................... 29
4.2.2 Dielectric Loss Factor ................................................................................. 30
4.3 DC Breakdown Field Strength .................................................................................. 30
4.4 Brief Summary ........................................................................................................... 32
5 Research on Trap Distribution of Modified Polyimide ........................................... 34
5.1 Mechanism of TSDC Experiment ............................................................................ 34
5.2 Test Equipment and Procedure ................................................................................. 35
5.2.1 Test Equipment ........................................................................................... 35
5.2.2 Test Procedure ............................................................................................. 35
5.3 Results and Analysis of Trap Energy Level Distribution of Modified Polyimide 36
5.3.1 Test Result and Data Process ..................................................................... 36
5.3.2 Trap Energy Level Distribution ................................................................. 38
5.4 Brief Summary ........................................................................................................... 41
6 Research on Conductance-Temperature Characteristic of Modified Polyimide ... 42
6.1 Test Platform of Nonlinear-Conductance Expriment .............................................. 42
6.2 Influence of Concentration of ZnO .......................................................................... 43
6.3 Influence of Grain Size of ZnO ................................................................................ 45
ACKNOWLEDGEMENTS
3
6.4 Influence of Temperature .......................................................................................... 45
6.5 Research on Nonlinear Conductivity Mechanism ................................................... 46
6.5.1 Mechanism of Nonlinear Conductivity at 25°C ........................................ 47
6.5.2 Mechanism of Nonlinear Conductivity at Different Temperatures ......... 52
6.6 Brief Summary ........................................................................................................... 55
7 Conclusion and Perspectives..................................................................................... 57
7.1 Conclusions ................................................................................................................ 57
7.2 Outlooks ..................................................................................................................... 58
Bibliografia ............................................................................................................................... 59
Author’s Publication ................................................................................................................ 61
ACKNOWLEDGEMENTS
4
Acknowledgements
There are people whom I would like to acknowledge, for their assistance and support during
my studies in Politecnico di Milano. I would like to thank all the wonderful teachers,
colleagues, family and friends whom I have been fortunate to interact with during my
lifetime.
I would like to take this opportunity to express my sincere gratitude and appreciation to my
supervisor professor Giovani Dotelli for his countless efforts in guiding and encouraging me
throughout my studies and work. His friendly attitude has been a very strong support for me
to work with him.
I am so grateful to him. I am also thankful to doctor Saverio Latorrata, for his valuable
advices in discussions, which was a plus point during my research. Without doubt, all these
discussions together guided me into a bright way to handle the research and studies.
Also I would like to thank my colleagues in the university, for their effortless helps, valuable
advices and discussions. We had great and unforgettable times during all these years.
The last but not least, I am so thankful to my family whom they have been a continuous
source of encouragements and supports in all directions during my life.
INDICE DELLE FIGURE
5
Indice Delle Figure
Figure 1: Charging Characteristics under Different Grounding Mode.[11] ......................... 11
Figure 2: Schottky Effect. ........................................................................................................ 15
Figure 3: Poole-Frankel Effect. ............................................................................................... 15
Figure 4: Relationship between SCLC and Electric Field ..................................................... 17
Figure 5: Conducting Charactisters of Modified ZnO.[23] ................................................... 20
Figure 6: Conducting Charactisters of Modified SiC.[23] .................................................... 21
Figure 7: Chemical Structure of Polyimide ............................................................................ 22
Figure 8: Chemical Structure of KH550 ................................................................................. 24
Figure 9: Dispersing Device[14] ............................................................................................. 26
Figure 10: Procedure of Compressing Molding ..................................................................... 27
Figure 11: Volumn Resistivity of Modified Polyimide ......................................................... 28
Figure 12: Relative Permittivity of ZnO/PI ............................................................................ 29
Figure 13: Dielectric Loss Factor of ZnO/PI .......................................................................... 30
Figure 14: Weibull Distribution of Breakdown Field ............................................................ 32
Figure 15: Circuit of TSDC Test ............................................................................................. 35
Figure 16: TSDC Current versus Temperature ...................................................................... 36
Figure 17: Trap Level Distribution ......................................................................................... 38
Figure 18: Trap Level Distribution of μm ZnO/PI ................................................................. 39
Figure 19: Trap Level Distribution of nm ZnO/PI ................................................................. 40
Figure 20: Conducting Test Platform...................................................................................... 42
Figure 21: Conductivity Test Result at 25°C ......................................................................... 43
Figure 22: Nonlinearity Comparison of μm and nm ZnO/PI (25°C) .................................... 45
Figure 23: Conductivity Test Result at Different Temperatures ........................................... 46
Figure 24: Nonlinear Conductivity Curve .............................................................................. 47
Figure 25: Schottky Effect Linear Fit at 25°C ....................................................................... 48
Figure 26: Tunneling Effect Linear Fit at 25°C ..................................................................... 48
Figure 27: SCLC Linear Fit at 25°C ....................................................................................... 50
Figure 28: Schottky Effect Linear Fit at Different Temperatures ......................................... 52
Figure 29: Tunneling Effect Linear Fit at Different Temperatures....................................... 53
Figure 30: SCLC Linear Fit at Different Temperatures ........................................................ 54
INDICE DELLE TABELLE
6
Indice Delle Tabelle
Table 1: Parameters of Polyimide Molding Powder ............................................................. 23
Table 2: Parameters of ZnO Power ........................................................................................ 24
Table 3: Parameters of KH550 ............................................................................................... 25
Table 4: Volumn Resistivityof Modified Polyimide............................................................. 28
Table 5: Breakdown Field of ZnO/PI ................................................................................... 31
Table 6: Shape Parameters and Average Breakdown Field Strength .................................. 32
Table 7: Trap Depth and Average Density of μm ZnO/PI ................................................... 39
Table 8: Trap Depth and Average Density of nm ZnO/PI.................................................... 40
Table 9: Threshold Field of Each Sample at 25°C................................................................ 44
Table 10 Threshold Fiels at Different Temperatures ............................................................ 46
Table 11: Transient Field of SCLC at 25°C .......................................................................... 51
Table 12: Transient Field of SCLC at Different Temperatures............................................ 55
ABSTRACT
7
Abstract
Polyimide is widely used in aerospace as insolation material because of its excellent
dielectrical, thermal, mechanical and physical properties. However, when running in the
space, spacecraft would be radiated by energized particles. Due to the high resistivity of
polyimide, it is difficult to release space charges once injected. Therefore, electrostastic
charges will accumulate inside polyimide, forming a local strong field. Also, those internal
charges will be radiated by infrared, ultraviolet, X-ray or visible light, thus give rise to pulse
discharge, which is known as deep dielectric charging. This phenomenon will threaten the
safety operation of spacecraft.
One way to resist the injection of high energy particles is building metal electrostatic
shielding, but it may weigh more and cost more. Researches show that the most effect way
to eliminate deep dielectric charging is by modifying the dielectrics with inorganic metal
oxides so that the conductivity of modified dielectrics will increase fast as the electrostatic
field increases. In this way the internal charges will be transferred in time and impulsive
discharge will be avoild. After that, the material will come back to high insolation state. This
is the though of nonlinear conductivity. Nonlinear conductivity modification of dielectric
material for spacecraft may become the main solution of deep dielectric charging.
In this thesis polyimide was modified by ZnO, and its nonlinear conductivity mechanism
was studied. First, dielectric tests were carried out to check whether the modified polyimide
could still be a good dielectric. Then, distribution of trap energy level was calculated after
the thermally stimulated discharge current (TSDC) tests to know the information of trap
energy level and trap density. At last, conducing tests in rising electric field at different
temperatures were carried out, and the relationship between current density and field strength
was discussed. Finally, nonlinear conductivity mechanism was concluded combing the result
of trap energy level distribution.
KEYWORDS: Polyimide, ZnO, modification, nonlinear conductivity, trap distribution.
ABSTRACT
8
Abstract (Italiano)
Il poliimmide è ampiamente utilizzato in ambito aerospaziale come materiale di isolamento
a causa delle sue eccellenti proprietà dielettriche, termiche, meccaniche e fisiche. Tuttavia,
quando operano nello spazio, le navi spaziali potrebbero essere irradiate da particelle
energizzate. A causa dell'elevata resistività del poliimmide, è difficile che questo rilasci le
cariche spaziali una volta iniettate. Pertanto, le cariche elettrostatiche si accumuleranno
all'interno del poliimmide, formando un forte campo elettrico locale. Questo fenomeno
minaccia la sicurezza delle navi spaziali.
Un modo per resistere all'iniezione di particelle ad alta energia è la costruzione di
schermature elettrostatiche metalliche, che, però, comporta più peso e costi più elevati. Le
ricerche dimostrano che il modo più efficace per eliminare la carica dielettrica profonda è
quello di modificare il materiale dielettrico con ossidi metallici inorganici in modo che la
conducibilità del dielettrico risultante aumenti velocemente all’aumentare del campo
elettrostatico. In questo modo le cariche interne saranno trasferite in tempo evitando le
scariche impulsive. In seguito, il materiale tornerà al suo stato originale con un alto livello
di insolazione. Questo è il concetto della conduttività non lineare. La modifica della
conducibilità non lineare del materiale dielettrico potrebbe diventare la soluzione principale
per le cariche dielettriche profonde nell’ambito delle navi spaziali.
In questa tesi il poliimmide è stato modificato con ZnO e il comportamento della sua
conduttività non lineare è stato studiato. Innanzitutto sono state effettuate prove dielettriche
per verificare se il poliimmido modificato possa ancora essere un buon dielettrico. Quindi,
la distribuzione del livello di energia di trappola è stata calcolata dopo prove di corrente di
scarica stimolate termicamente (TSDC) per conoscere le informazioni del livello di energia
di trappola e della densità di trappola. Infine, sono stati condotti test di conduzione con
campo elettrico crescente a diverse temperature, e sono stati discussi i rapporti tra densità di
corrente e forza del campo. Infine, sono riportate conclusioni riguardo il meccanismo di
conducibilità non lineare combinandole con il risultato della distribuzione del livello di
energia di trappola.
KEYWORDS: Polyimide, ZnO, modifica, conduttività non lineare, distribuzione trappola.
INTRODUCTION
9
1 Introduction
1.1 Introduction of Deep Dielectric Charging
1.1.1 What is Deep Dielectric Charging
Deep dielectric charging is also known as internal dielectric charging. It is the process that
high power electron beam (100keV~10MeV) penetrates dielectric and deposites inside the
material. High power electrons usually spread outside the earth’s radiation belts (3~7RE)
and have strong penetrating power. Their energy varies from 0.1MeV to 10MeV. They can
not only penetrate dielectrics outside spacecrafts, but also penetrate shielding materials and
deposit in the dielectrics inside spacecrafts, such as printed circuit boards and insulation
sleeves of cables. Since dielectrics generally have high resistivity, it is difficult for electrons
to move, which will lead to high potential inside dielectrics, up to ten thousands volts. Those
electrostatic charges accumulated in surface or internal will be radiated by infrared,
ultraviolet, X-ray or visible light, thus trigger pulse discharge[1,2]. Pulse discharge current
and high-frequency electromagnet wave coming along with the discharge will disturbe
accurate electronic instruments and thus affect the normal operation of the system. In severe
cases, sensitive electronic components or organic dielectrics will be breakdown.
It goes through three steps from high power electron penetration to discharge, including
injection and deposition of electrons, formation of internal field and discharge of dielectric.
When the rate of charge injection exceeds charge release, charges will built up inside
dielectrics, and space electric field will be generated gradually. When the field strength
exceeds breakdown strength, discharge happens. Since potential difference is usually quite
large under this situation, discharge happens in the form of pulse discharge. According to
statistics, 84.76% of space environment fault is caused by charged particles.
Domestic and overseas scholars have made research on the quantitative evaluation of deep
dielectric charging in spacecraft. The cumulative injection amount of high-energy electron
above 2MeV (which can penetrate 2.5mm thick Aluminum shelding layer) is usually used
as the macroscopic evaluation of the danger of deep dielectric charging and discharging.
Friderickson put forward this evalution system that the cumulative injection of electronic
above 2MeV, that is, F>2MeV=1.8×1010cm-2 is the threshold of deep dielectric charging and
discharging[3].
INTRODUCTION
10
1.1.2 How to Solve Deep Dielectric Charging
Since the seventies of last century, research on the fault of spacecraft caused by charging
and its protection technology have been paid more and more attenetion.
The tradition method of space charged protection is by generating electron emission or ion
emission to decreasing potential, while the method of deep charge protection is by using
Aluminium shielding. The thickness of Aluminium shielding should be no more than 3mm
considering the weight and cost. However, charging problem still cannnot be completely
avoid. Higher energy electrons (>3MeV) still can penetrate Aluminium shielding and
accumulate in the dielectric in the cabin.
Another way is by increasing the leakage rate of internal deposit charge, but this only suits
for dielectrics used not for insulating purpose, such as the thermal control layer made of
semi-conductive polyimide composite at the surface of spacecraft.
From the last section we know that the high resistivity of dielelctrics is the main reason of
the difficulty of the static charges leakage. So under the promise of satisfying the basic
electrical properities, mechanical properities and chemical properities, properly increasing
the conductivity of dielectric can to a certain extent inhibit the accumulation of static charges,
but it will increase energy loss of power electronic system. Researches show that dielectrics
with nonlinear conductivity may fundamentally solve the deep dielectric charging
problem[4]. That is, under normal condition, the maiterial has high resistance. But when the
deep static field exceeds a certain value, the conductivity will be in a rising transient state
and low energy conductance discharge will occur. After that, the material will be in high
insulation state again. This method may be the mainstream method to defend deep charge in
space dielectrics.
1.1.3 Influence of Grounding Condition
In spacecraft, there are four ways to ground dielectrics: ungrounding, radiant side grounding,
back side grounding and two sides grounding.
INTRODUCTION
11
(a) ungrounding (b) radiant side grounding (c) back side grounding (d) two sides grounding
Figure 1: Charging Characteristics under Different Grounding Mode[5].
Where:
(a) In the case of ungrounding, electrostatic charges will continue rising as high energy space
charges injecting into the material, in spite of how much energy the electron beam has. In
this case, if one-point grounding happens, it will immediately lead to electronic trees, which
will damage the material.
(b) In the case of radiant side grounding, electrostatic charges are more easily to be released
because radiation induced conductivity is several orders of magnitude higher than that of
intrinsic conductivity. No matter how much electron energy is, charge accumulation will not
happen, and the media is relatively safe.
(c) There are two conditions in the case of back side grounding. One is that the energy is
sufficient for electrons to penetrate the media. The other one is that electrons could not get
through the media. In the latter condition, electrostatic charges will accumulate and
electrostatic field will increase, finally the media will be breakdown.
(d)In the case of two sides grounding, there will be no charge accumulation.
Dielectrics in spacecraft usually use the back side grounding method. There is risk of
breakdown. If the material has the ability of releasing electrostatic charges automatically,
just like in the case of radiant side grounding, deep dielectric charging problem can be solved.
Therefore, it is important to study the deep charge and discharge mechanism of dielectrics
used in the space environment and materials with nonlinear conductivity which can release
charges automatically. It is important not only for the reliable operation of spacecraft, but
also for the power electronics system and design of spacecraft life.
INTRODUCTION
12
1.2 Contents of work
The main contents of this paper are as follows:
(1) Polymer modification technology was studied and ZnO modified polyimide samples
with different ZnO sizes and concentrations were prepared.
(2) The dielectric properties were analyzed systematically to see whether basic insulation
requirements could be met after modification.
(3) The depolarization current analysis of each sample was carried out, and the trap density
distribution was calculated according to the experimental results. The effects of nano
modification and micron modification on the trap distribution were studied.
(4) The conduction test system which can be used at variable temperatures was designed to
study the influence of the concentration of modifier, the particle size of the modifier and
the ambient temperature on the nonlinear conductivity of the modified polyimide. The
experimental results were processed to obtain the conductivity model satisfying the
nonlinear conductivity, and the nonlinear conductivity mechanism was explained
together with the result of trap energy distribution.
CAPITOLO 2
13
2 Theory Background
2.1 Conductivity in Space
Conductivity is the core parameter to measure the electrostatic properties of dielectric
materials. Under space environment, the conductivity is consisted of:
0= + +E (1)
Where:
σ0 is the conductivity in normal condition, and it is also named as “dark conductivity”. Its
value depends on the number of carriers and mobility. Since the carrier concentration of
good dielectric is very small (usually under 10-20/cm3), the intrinsic conductivity is very
small. Conductivity of polar materials is 1 or 2 orders of magnitude larger than that of non-
polar materials due to its asymmetric molecular structure. Commonly the conductivity of
good dielectric is around 10-18/Ωcm.
σΔ is induced conductivity. In radiation environment, dielectric material will have the so-
called radiation induced conductivit (RIC), which is a few orders of magnitude higher than
the intrinsic conductivity. Induced conductivity can be divided into transient induced
conductivity and permanent induced conductivity. The former one is caused by the excitation
of some electrons from valence band to conduction band by high-energy particles, and it will
disappear when the high-energy ray disappears. The latter one is permanent, and it is caused
by raditon aging.
σE is nonlinear conductivity, also known as electric field induced conductivity. It usually
appears before breakdown, and accompanied by damage to material.
The conductivity mechanism of nonlinear modified dielectric materials is still not completed.
Although some scholars have proposed relevant models and theoretical analysis, a lot of
repetitive experiments are needed to form a pervasive and systematic theory, together with
the consideration of the influence of modifier, modifier concentration, temperature, pressure
and other factors to make the theoretical research more abundant and perfect.
CAPITOLO 2
14
2.2 Classic Conducting Theory
In general, the interpretation of the conductive mechanism of polymers is usually based on
several classical dielectric physics theories, including hopping conduction, Schottky effect,
Poole-Frankel effect, tunneling effect and space charge limited current effect. The
followings are brief description of these effects.
2.2.1 Hopping Conduction
Hopping conduction can be divided into ion hopping conduction and electron hopping
conduction. Most of the insulating polymers have a structure in which the crystal phase and
the amorphous phase coexist, and the electrons in the amorphous phase region need to
overcome a potential barrier u0 when migrating from one conduction band to the adjacent
conduction band. This migration tends to rely on the thermo-electronic transition, that is,
under the action of thermal vibration, the electrons on the local conduction band transit to
the adjacent conduction band to form the electronic hopping conduction. The ion hopping
conduction is similar.
The relationship between current density formed by hopping conduction and the electric
field can be expressed by the following equation.
0 0exp sinh3 2k
qn u q Ej v
kT T
(2)
Where: q——carrier charge/C; n0——carrier concentration/cm-3; v——vibration
frequency/Hz; δ——average hopping distance/m; u0——average barrier to be overcome in
the transition/eV; k——Boltzmann constant; T——absolute temperature/K; E——electric
field strength/V·m-1.
It can be seen that current density j versus electric field E shows a hyperbolic sine function.
2.2.2 Schottky Effect
When there is an external electric field E, the potential barrier of the cathode interface
decreases, so the potential barrier of the electron to escape cathode reduces, and the
thermoelectron emission current increases. This phenomenon is Schottky effect, as shown
in Figure 2.
CAPITOLO 2
15
Figure 2: Schottky Effect.
The relationship between thermoelectric emission current and applied electric field is:
2A exp s DEj T
kT
(3)
Where: A——Richardson-Dushman constant; βs —— Schottky constant; D ——Actual
work function.
2.2.3 Poole-Frankel Effect
Inside the medium there is also a phenomenon similar to the Schottky effect. The
phenomenon that potential barrier reduces under external electric field and electronic
conduction current increases is known as the Poole-Frank effect. Figure 3 shows the effect
of the electric field on the internal barrier of the medium.
Figure 3: Poole-Frankel Effect.
The relationship between the conductivity and the electric field is as follows:
CAPITOLO 2
16
0 exp F E
kT
(4)
Where: 0——intrinsic conductivity /S·m-1; βF——Poole-Frankel constant.
2.2.4 Tunneling Effect
Under strong electric field, when the barrier thickness is very thin, and the electronic energy
is very close to the barrier height, the electrons may pass directly through the barrier without
lossing energy. This phenomenon is called tunneling effect. The relationship between the
current density and electric field strength can be expressed as:
2 expB
j AEE
(5)
Where: A,B—— constants.
2.2.5 Space Charge Limited Current Effect
When carriers migrate in the medium, they may be bound in the local region, forming space
charges, which in turn causes the space charge limited current.
Under strong electric field, electron current is the main conducting current. Current injected
from electrode Ic continuous with electron current in the dielectric Ib. At steady state, the two
currents are equal. If Ic≠Ib, electrons will accumulate inside the dielectric. In the case of Ic<Ib,
positive space charges are formed near cathode to increase the cathode injection current, so
the injection current Ic will increase to equal Ib.; in the case of Ic>Ib, negative space charges
are formed near cathode to reduce cathode electric field and Ic. At the same time, space
charge limited current Is occurs until Ic=Ib+Is. The relationship between space charge limited
current density and electrode voltage can be expressed by Kader's law [6].
2
0s 3
9
8
r Uj
d
(6)
In the equation: εr——relative permittivity; ε0——vacuum dielectric constant /F·m-1; μ——
carrier mobility /cm2·(V·s-1)-1; U——voltage/V; d——electrode distance/mm.
Take the average field strength Ea=U/d, the following expression is derived:
CAPITOLO 2
17
2
09
8
r aEj
d
(7)
It is thus clear that space charge limited current density is proportional to the square of the
electric field strength, and inversely proportional to the thickness. This relationship is called
the Mott-Guinai relationship[7].
Theoretically the relationship between sclc current and voltage is shown in Figure 4. In the
figure, segment a is the ohmic region, the slope of which is 1; segment b and c are sclc region
with a slope of 2. If there are electron traps in the dielectric and are not filled, electrons may
be trapped, so a portion of the electrons injected from the electrodes are trapped as space
charges, while others are conductive. In this case, the input charges are more than the output
charges. As electric field increases, charges injected from cathode will increase. When traps
are all filled, input charges equal to output charges and the space charge limited current will
increase sharply. At this time, segment a transfers to segment b. When electric field increases
further more, charge energy will be greater than trap depth, and trapped charges will detrap
and all involve in the conduction. At this time, input charges are less than output charges,
and current density increases rapidly, and now it transfers to segment c. EiT is the transition
field from ohmic region to sclc region, Eab is the field when traps are all filled, and ET is the
field when charges are all detrapped.
Figure 4: Relationship between SCLC and Electric field
From the expressions of above conduction mechanism, we can see that conductivity and
electric field have a certain relationship. At present, research on the mechanism of polymer
conductivity is first to measure the data through experiments, then use mathematical tools to
CAPITOLO 2
18
process the data and compare them with these theories. Finally the conduction mechanism
can be determined.
For the polymer-based modified materials, since the polymer itself is short-range orderly
and remote-range disordered with composite structure of crystal phase and amorphous phase
coexistence, it is even more complex when modified by inorganic filler. The movement of
carriers is unavoidably affected. Therefore, the conductivity of modified material is complex
and may contain a variety of conductive mechanisms.
CAPITOLO 3
19
3 Modification of Polyimide
3.1 Develop of Modification Technology
Generally, nonlinear conductivity modification can be realized by adding inorganic
compound with high conductivity. The modifier is typically semi-conductive inorganic
metal oxide in size of micron or nanometer. The particle size of micron filler is usually in
the range of 1 to 10 μm. The particle size of nanofiller is usually in the range of 1 to 100 nm.
Nano filler has more properties such as surface effect, small size effect, Kubo effect and
macroscopic quantum tunneling effect, which can deepen the trap depth in polymer, suppress
space charge, inhibit the development of electric trees[8].
There are two key indicators to evaluate nonlinear conductivity: the threshold electric field
and the nonlinear steepness[9].
1) threshold field
The threshold field is the equilibrium field of charge accumulation and release. If the
threshold field is low, conductivity can begin to increase when the local field strength
exceeds the threshold field in a low value. In this case, electrostatic charges can be released
early. And once the electrostatic field stops increasing, the material returned to stable state.
For most materials, the threshold field should be much higher than working field and much
lower than breakdown field.
2) nonlinear steepness
The nonlinear steepness has an important effect on the charge release process. If the slope
of the surge after threshold field is steep, charges can be released fast but strong pulse
discharge may occur. If the slope of the surge after threshold field is gentle, charges are
released slowly and it is on the contrary to the original intention. So the increase speed of
conductivity should be appropriate. It should be at a fast rate to release charges quickly while
at the same time not forming any dangerous pulse discharge. At present, there has not a
quantitative index to settle the nonlinear steepness.
Study on the nonlinear conductivity modified materials is mainly concentrated in the
atmospheric environment and commonly used polymers, and it has already been used for the
CAPITOLO 3
20
terminal anti-corona structure of cable and insulation sleeve of extra high voltage. For the
mechanism study, there has not been an entire theory system. There are many studies on the
influence of the modifier to the conductivity but few studies on the temperature. Additionally,
a lot of repetitive experiments are still needed in the theoretical study to obtain reliable
conclusions and also to enrich dielectric physics theory. At present, the commonly studied
polymers are epoxy resin, polyimide, polyethylene, low density polyethylene and rubber,
and the commonly used modifiers are ZnO, SiC and so on. The following lists are researches
on the nonlinear modification.
In the year of 2003, K.P.Donnelly and B.R.Varlow[10] added ZnO (particle size: 6μm) to
polyester resin and prepared modified samples (thickness:1.2-2.0mm). It was found that
when the concentration of ZnO was higher than 30wt%, the current density increased sharply
with the increase of field strength, and no longer met Ohm's law, as shown in Figure 5.
Likewise, when the modifier was SiC, the modified samples also showed a similar
performance. When the concentration of SiC was 60 wt% or more, the threshold field
strength was as low as 0.23 kV/mm, as shown in Figure 6. If compared with samples
modified by Al2O3, it was found that the nonlinear conductivity was caused by ZnO and SiC,
and not any modified materials had nonlinear conductance.
Figure 5: Conducting Charactisters of Modified ZnO[10].
CAPITOLO 3
21
Figure 6: Conducting Charactisters of Modified SiC[10].
In the year of 2009, Guo Wenmin and Han Baozhong[11] of Harbin Polytechnic University
studied the main factors influencing the conductance of ZnO/LDEP. The results showed that
the nonlinearity of ZnO/LDPE was impoved as the increasing of ZnO concentration,
increasing of the temperature and reducing of the pressure. Further more, The different grain
structure of ZnO caused by different production processes resulted in the differences of the
nonlinear conductivity characteristics.
In the year of 2010, Guo Wenmin[12] of Harbin Polytechnic University did the conducting
test for ZnO/PE (ZnO concentration:11.62 vol%) at 293K, 308K and 323K respectively. It
was found that conductivity incerased as temperature increased. This indicated that there
existed thermal hopping process in the conductivity mechanism. At the same time, threshold
field tended to move to lower values as temperature incerasing in the log-log coordinate.
In the year of 2016, Liu Changyang[13] of Xi'an Jiaotong University prepared 2μm and
10μm ZnO modified epoxy resin and did conducting tests and space charge tests. It was
found that both modified samples had nonlinear conductivity characteristics, And space
charges dissipated as electric field increases. The relationship between the two tests was
analyzed, and it was considered that holes neutralized negative space charges in the interface
energy band, which decreaesd the space charge density. And the decrease of high energy
bandwidth made the iincrease of tunnel current.
At present, there have been mature commercial products of modified polymers. For example,
medium-voltage power cables use dielectrics modified by ZnO voltage-sensitive ceramic
CAPITOLO 3
22
powders for cable terminal insulation, and it has a good inhibitory effect on the formation of
electric trees. Further more, due to the stable nonlinear barrier interface of ZnO voltage-
sensitive ceramic particles, the overall performance of the modified material is greatly
improved[14].
3.2 Main Features of Raw Materials
3.2.1 Polyimide
Polyimide is a kind of organic polymer which refers to a class of polymers containing an
imide ring (CO-NH-CO-) in the main chain. Its structure is shown in Figure 7.
Figure 7: Chemical Structure of Polyimide
Polyimide is widely used in aerospace, machinery, electronics and other high-tech modern
areas[15,16] because of its good overall properties, especially thermostable performance.
Thermal decomposition temperature of polyimide is up to 500-600°C, and the range of
working temperature is from -200°C to 300°C. It is one of the polymers which have highest
thermal stability so far. At the same time, polyimide is an excellent dielectric material. Its
dielectric constant is about 3.4 at 1kHz, and dielectric loss is 4×10-3 to 7×10-3, belonging to
F to H class of insulation material. Additionally, polyimide has excellent mechanical
properties and anti-radiation properties, which makes it an important material in the field of
aerospace.
In this thesis, SKPI-MS30 (general type) of thermoplastic PI molding powder was used,
produced by Changzhou Shangke Special Polymer Co., Ltd. Its median particle size is about
50μm. Parts of the parameters are shown in Table 1. The reason for choosing molding
powder is that it is easy to mix mozxv difier powder with it, which is suitable for future
industrial production.
N
C
C
O
O
R
C
C
O
O
R'N
n
CAPITOLO 3
23
Table 1: Parameters of Polyimide Molding Powder
Iteams Properties
Appearance Light yellow power
Density/kg·m-3 1400
Water absorption(25°C, 24h)/% ≤0.6
Glass transition temperatureTg/°C 250-260
Weight loss 5 wt% decomposition temperatureTd5/°C 530
3.2.2 ZnO
ZnO is a kind of inorganic oxide. Resistivity of unsintered ZnO is very small, around 100
Ω·cm. It is a kind of semiconductor material, and does not have the properties of nonlinear
conductivity itself. The structure of ZnO crystal is hexagonal wurtzite, the lattice constant a
= 0.325nm, c = 0.521nm.
In this work, nano ZnO and micron ZnO were used to modify polyimide. Particle size of
Nano ZnO is between 1-100nm, and it is a multi-functional inorganic material with special
properties in the light, electricity, magnetic, sensitive and so on. Nowadays, it is widely used
in rubber, ceramics, power electronics, coatings and other industrial fields. Compared with
ordinary ZnO, its surface structure and crystal structure change, and has special properties
such as surface effect, volume effect, quantum size effect and macroscopic tunneling effect.
However, because of its large specific surface area and specific surface energy, it is easy to
agglomerate itself, and it is not easy to disperse evenly in organic medium.
The ZnO powder used in this paper is provided by Shanghai Paddy Field Materials
Technology Co., Ltd. Table 2 is its parameters, in which the nano ZnO is represented by
MON and the micron ZnO is represented by MOW.
CAPITOLO 3
24
Table 2: Parameters of ZnO Power
Iteam
Properties
MOW MON
Appearance White powder White powder
Average particle size/nm 1000 50
Density/kg·m-3 5600 5600
Volume density/kg·m-3 500 300-450
Specific surface area/m2·g-1 40 ≥100
Resistivity/Ω·cm 100 100
3.2.3 Coupling Agent
KH550 is a kind of silane coupling agent, and its molecular formula is
NH2(CH2)3Si(OC2H5)3. Chemical structure is shown in Figure 8.
Figure 8: Chemical Structure of KH550
As the surface properties of inorganic particles and organic polymer are of great difference,
the compatibility of this two materials is very poor. ZnO particles have a large number of
defects and hanging bonds on the surface with high degree of unsaturation and large surface
energy, which leads to strong chemical reactivity. For nano ZnO, due to its small particle
size and large specific surface area, it has large surface energy and surface bonding energy,
which makes it easily to agglomerate. Therefore, it is necessary to do surface treatment as
pre-processing to increase the interfacial bonding force between the modifier and the
polyimide matrix. KH550 is an alkaline coupling agent, and it is commonly used to improve
the wettability and dispersibility of the filler added into the polymer.
CAPITOLO 3
25
In this work, KH550 is provided by Jiangsu Dandelion New Materials Co., Ltd. Part of the
parameters are shown in Table 3. The amount of silane coupling agent is generally 0.1% to
3.0% of the mass of inorganic particles. For nano ZnO, since it has large specific surface
area, more coupling agent is needed to insure that all particles are well coated. Therefore,
the amount of KH550 is 3 % of nano ZnO. For micron ZnO, the amount of KH550 is 1%.
Extra coupling agent is removed by filtration.
Table 3: Parameters of KH550
Iteams Properties
Appearance Colorless liquid
Density/kg·m-3 946
Boiling point /°C 217
Refractive index/nD25 1.420
3.3 Modification Process
The preparation process of modified polyimide is mainly divided into two steps. The first
step is to mix the polyimide powder and ZnO powder. The second step is to mold the mixed
powder into plate samples.
3.3.1 Powder Mixing
Modifier and polyimide were dispersed in liquid medium, and the mixture was thoroughly
stirred and then filtered and then dried. Figure 9 is the diagram of the dispersing device.
There is a rotor driven by motor to rotate at high speed, and the dispersing tool is rotated at
high speed in the solution. The agglomerated particles in the solution are dispersed under
strong shear forces and severe mechanical forces. At the same time, the solution is subjected
to ultrasonic shaking. Ultrasound acts on liquid molecules so that microbubbles appear in
the liquid and develop into cavitation bubbles. Cavitation bubbles vibrate violently, finally
collapse into high speed micro jet, which can exert a strong impact force on the aggregations.
This process makes the agglomerations collaps into tiny aggregations or even a single
particle. In this way, homodispersion of the modifier and matrix powder can be achieved.
CAPITOLO 3
26
Figure 9: Dispersing Device[14]
The complete procedure is as follows:
1) The absolute ethanol and deionized water were mixed at a ratio of 95:5. A certain
amount of modifier powder and polyimide powder were added into the solution together
with an appropriate amount of KH550.
2) The above mixed solution was placed in an ultrasonic shaker to do the ultrasonic
oscillation. At the same time an agitator was stirring the solution at the speed of
200r/min for 7 mins.
3) The dispersed solution was put into a vacuum oven at the temperature of 120°C for 24
hours to dry out the solution. After drying, powders were grinded and sieved to be well
prepared.
3.3.2 Compression Molding
Compression molding is the most commonly used technology in composite production. This
technique has been developped for a long time, and has the advantages of small loss of raw
material, small internal stress, low equipment cost, high productivity and so on.
If the mixed powder is directly pressed at high temperature and pressure, it is easily to get
bubbles inside samples, and the thickness of the samples can not be guaranteed. So firstly
the mixed powder should be pressed at room temperature to have a certain shape and
thickness. This procedure is called ”cold moulding”. The powder was placed in a cylindrical
mold with a diameter of 70mm. The molding pressure was set as 10MPa, and the molding
was kept for 3s each time. The total pressing time was 5. After cold moulding, pressed
powder samples with a diameter of 76 mm and a thickness of 1 mm were prepared.
CAPITOLO 3
27
Then the pressed powder samples were put into the flat vulcanizing machine. The following
procedure in Figure 10 was set.
Figure 10: Procedure of Compressing Molding
After the above process, demould the samples. The ZnO modified polyimide samples are
translucent plate-like samples with a diameter of 80 mm and a thickness of 0.8 mm, and the
transparency was reduced as the amount of ZnO was increased.
In this work, pure polyimide samples, nano ZnO modified polyimide samples and micron
ZnO polyimide samples were prepared, wherein the addition amount of ZnO 1 wt%, 2 wt%,
3 wt%, 5 wt% respectively. For simplicity, each sample is represented by the following name:
pure PI, 1 wt% μm, 2 wt% μm, 3 wt% μm, 5 wt% μm, 1 wt% nm, 2 wt% nm, 3 wt% nm, 5
wt% nm.
CAPITOLO 4
28
4 Basic Dielectric Performance
Modified polyimide samples still need to meet basic electrical standards as an insulating
material. The volume resistivity, relative dielectric constant, dielectric loss factor,
breakdown field strength tests were carried out to see if the insulation requirements still
satisfied.
4.1 Volumn Resistivity
6517 electrometer and 8009 resistivity test cartridge were used to run the volume resistivity
test. The test conditions were: 20°C, 36% relative humidity, 800V applied voltage. The
results are shown in Table 4 and Figure 11.
Table 4: Volumn Resistivity of Modified Polyimide
Volume resistivity/1017Ω·cm 0(Pure PI) 1wt% 2wt% 3wt% 5wt%
μm ZnO/PI 3.10 2.40 1.41 1.67 1.41
nm ZnO/PI 3.10 2.04 1.06 1.50 0.94
Figure 11: Volumn Resistivity of Modified Polyimide
The resistivity of modified samples are lower than that of pure PI samples, but still on the
order of 1017Ω·cm, which are quite high values as insulating material. The resistivity of
micron modified samples is larger than that of nano modified samples. And with the increase
of ZnO additives, the resistivity decreases, and the decreasing trends of micron and nano
modified samples are the same. The reason maybe that at the same amount of ZnO, the
CAPITOLO 4
29
number of nano particles per unit volume is much larger than that of micro particles, and it
is easy to form the nano "bridge", along which the current can flow, so nano modified
samples have smaller resistivity.
4.2 Dielectric Spectrum
Concept 80 Wideband Dielectric Spectrum Tester was used to test dielectric frequency
spectrum from10-2 Hz to 107 Hz at room temperature. Test results are as followed.
4.2.1 Relative Permittivity
Figure 12: Relative Permittivity of ZnO/PI
The relative permittivity of each sample decreases monotonically with the increase of
frequency, and the tendency is quite flat. It means that modifier does not change the
polarization of the material, which is relaxation polarization between two phases. The
relative permittivity of micron modified samples is between 3.6 and 4.3, and between 3.6
and 5.1 of nano modified samples. Besides, the variation trend of the two modifiers with the
same particle size is the same as the change of ZnO mass fraction. In general, the variation
trend of the relative permittivity is 5 wt%> 3 wt% > Pure PI> 2 wt%> 1 wt%.
Similar results are also found in others' studies[17] that low amount of modifier reduces the
relative permittivity. It is believed that dopant improves the interaction between molecular
chains of the polymer, which leads to molecular chain motion being blocked, so the
polarization rate decreases as a whole. With the increase of the doping mass, the relative
permittivity increases because that ZnO has large relative permittivity. From the testing
rusults we can see that relative permittivity of 5wt% nm is much higher than that of other
samples. It is considered that interfaces of 5wt% nm are of great amount and some of them
CAPITOLO 4
30
may overlap to form conductive channels. Also this consideration proves the "nano bridge"
proposed in Section 2.2.1. When ZnO is in micron size, a similar "mesh bridge" will not
appear, so the effect of modification on relative permittivity is not as pronounced as the
former.
4.2.2 Dielectric Loss Factor
Figure 13: Dielectric Loss Factor of ZnO/PI
At low frequencies, dielectric loss is contributed by the conductance loss. At this time, pure
PI has larger conductivity than modified samples, so its dielectric loss factor is the largest.
With the increase of frequency, dielectric loss of modified samples is larger than that of pure
PI, reaching the peak at about 100 kH, which is the relaxation region of polarization. In the
relaxed area, dielectric loss of micron modified samples is on the order of 0.001, and the
peak loss is about 0.01. Dielectric loss of nano modified samples is higher, especially 5wt%
nm, where the peak is about 0.03 while the remaining samples are below 0.02. One possible
reason is that 5wt% nm sample has more ZnO and the number of polarizations is large, so
the dielectric loss is large. What’s more, the formation of "nano bridge" makes this effect
more remarkable. The variation trend of the dielectric loss is 5 wt%>3 wt% >2 wt%>1 wt%,
which is similar to the trend of permittivity.
4.3 DC Breakdown Field Strength
The DC withstand voltage test was tested on pure PI, 1wt% μm, 3wt% μm, 5wt% μm, 1wt%
nm, 3wt% nm and 5wt% nm. Spherical electrode was used as test electrode. The
experimental data are collated and the field strength values are shown in Table 5.
CAPITOLO 4
31
Table 5: Breakdown Field of ZnO/PI
Iteam Pure PI
μm ZnO/PI nm ZnO/PI
1wt% 3wt% 5wt% 1wt% 3wt% 5wt%
Breakdown
field
kV/mm
144.80 114.50 140.04 97.35 100.66 87.23 67.15
150.30 163.02 142.00 111.41 109.26 104.35 70.98
157.98 163.16 145.21 114.76 119.28 105.32 81.14
165.23 164.23 157.38 117.88 123.06 106.70 83.87
167.18 164.59 160.84 118.29 123.17 115.93 84.74
168.53 177.10 166.64 129.63 126.81 123.48 85.98
175.27 181.11 177.91 130.38 150.41 124.09 88.15
186.16 140.69 98.47
142.94
Weibull statistical distribution method was used to process the data, which can clearly
express the dispersion degree and average breakdown field. X = lnEb was taken as the
independent variable Y=ln(-ln(1-F(Eb))) was taken as the dependent variable. Linear
regression result is shown in Figure 14. Slope m is the shape parameter, and the larger the
shape parameter, the smaller the dispersion of the breakdown data. Average breakdown field
strength was gotten as the corresponding field to F(Eb)=63.2%[18]. Shape parameters and
average breakdown field strengths of Weibull distribution are shown in Table 6
CAPITOLO 4
32
Figure 14: Weibull Distribution of Breakdown Field
Table 6: Shape Parameters and Average Breakdown Field Strength
Iteam Pure PI
μm ZnO/PI nm ZnO/PI
1wt% 3wt% 5wt% 1wt% 3wt% 5wt%
Shape paramete/m 14.64 10.88 11.14 9.13 7.78 9.01 8.77
Average breakdown
field/kV/mm 170.72 175.91 165.67 131.63 132.95 117.92 89.12
It can be seen that breakdown field of pure PI is the most concentrated, with average value
of 170.72 kV/mm. With the increase of ZnO, the dispersibility of the sample increases, and
the dispersion of nano modified sample is larger than that of micron modified sample at the
same ZnO concentration. Meanwhile, average breakdown strength of nano modified sample
is less than that of micron modified sample. It is found that 1wt% μm has a higher average
field strength than pure PI, probably due to the fact that little amount of ZnO are
homogeneous dispersed in the matrix and enhance the scattering effect of dipoles[19] DC
breakdown field strength of all the samples except 5wt%nm is above 110kV/mm, which is
quite a high value. The lowest breakdown strength of 5wt% nm further confirms the
formation of nano bridge in the matrix.
4.4 Brief Summary
Basic dielectric property tests showed that volume resistivity of modified samples decreased
slightly, but still high enough to ensure insulated. Low doping (1wt%, 2wt%) limited the
CAPITOLO 4
33
movement of molecular chain so that relative permittivity and dielectric loss decreased while
high doping (3wt%, 5wt%) increased the relative permittivity and dielectric loss. DC
breakdown field of modified samples was slightly lower, but still had strong electrical
strength. In the 5wt% nm sample, the nano bridge was formed due to the relatively large
doping amount, and the change of dielectric parameters is quite evident. In general,
dielectrical parameters of modified samples can still meet the insulation requirements.
CAPITOLO 5
34
5 Research on Trap Distribution of Modified Polyimide
The environment temperature of synchronous orbit spacecraft cycles for 24 hours. The
temperature of dark side can be -160°C, while the temperature of sunny side can be 200°C.
Since the dielectric material has very high resistivity under low temperature, the resistivity
can decrease several orders of magnitude when the temperature increases. So the
accumulation of injected charges happens in the dark side, while discharge happens in the
sunny side. Every time the change from dark side to sunny side will give rise to discharge.
Research on the thermally stimulated current of material can reveal the rules of accumulation
and release of electrostatic charge. In addition, it can also evaluate the trap energy level, trap
density and trap depth. In this chapter, the trap distribution of material was calculated by
Thermally Stimulated Depolarization Currents (TSDC) Test for further research on space
charge.
5.1 Mechanism of TSDC Experiment
In the year 1964 Bucci and Fieshci first proposed the theory of TSC to analyse dipole
polarization[20]. In their theory, thermally stimulated current was caused by structure
defects of ionic crystal, such as vacancy, dislocation and interstitial defect. Later, the basic
theory of thermal stimulation current was becoming mature, and the test technology
developped fast. It became widely used for dielectric material test. Thermally stimulated
current method can be divided into thermal stimulated polarization current method and
thermal stimulated depolarization current method. Usually TSC refers to thermal stimulated
depolarization current method. In this chapter, thermal stimulated depolarization current
method was used.
TSDC test is commonly used to study trap structure inside dielectric and storage and
transportation properity of space charge. It can observe the performance of sample’s
depolarization current as temperature increasing to know the changing process of internal
charges, therefore the trap properties can be known. This test can observe how the charged
particles “frozen” at nonequilibrium condition at low temperature comes back to the thermal
equilibrium state as temperature increasing. It is helpful to research on the trap structure and
space charge transportation property inside dielectrics.
CAPITOLO 5
35
During the test, first set the temperature and voltage to suitable values, and keep it for several
mins to make all polarizations happen. Then cool down the temperature to dozens of degree
below zero at the rate of 30°C/min and withdraw the voltage. After that, rise the temperature
to around 100°C at the rate of 2°C/min. At the beginning of the heating process,
depolarizations almost doesnot happen cause the temperature is too low, thus there is no
current in the external circuit. As temperature increases, the activity of impurity ions, dipoles
and molecular chain increases, and depolarizations occurs, and depolarization current can be
detected in the external circuit. When the depolarization process completes at a certain
temoerature, the current turns to zero again. The following is the test circuit.
Figure 15: Circuit of TSDC Test
5.2 Test Equipment and Procedure
5.2.1 Test Equipment
The test system includes temperature control system, a vacuum equipment, an electrometer
and a DC high voltage source. System parameters are: polarization voltage:0~1kV; current
range: 1 fA~20mA; temperature control: -160~300°C; heating rate: 0.01~30°C/min.
5.2.2 Test Procedure
Before the test, samples were short-circuited under 60°C for 12h to remove water and stray
charges. Then samples were sprinkled of gold powders to form a gold electrode with
diameter 30mm. After that, the sampple can be put into the test system.
First, sample was polarized under 70°C and 250V for 30min. Then cool the temperture fast
to -50°C by liquid nitrogen. At the same time, voltage was withdrawed. Then temperature
CAPITOLO 5
36
was increased at the rate of 2°C /min, until it reaching 100°C. Current induced in the external
circuit was recorded during the process.
5.3 Results and Analysis of Trap Energy Level
Distribution of Modified Polyimide
5.3.1 Test Result and Data Process
Figure 16 is the TSDC test result. During the heating process, β and ρ relaxation apperaed.
β relaxation at low temperature came from movement of micromolecule from side chain and
local movement of main chain, while ρ relaxation at peak current came from release of space
charge[8].
Figure 16: TSDC Current versus Temperature
To calculate trap parameters, the improved quasi-continuous distribution method was used.
In geneeal, energy level is considered to be discreted. But for polymer, because of its
complcated inner structure, low purity and large amount of interfaces caused by modifier,
environment near traps cannot be exactly the same, which leads to the non-discrete
distribution of energy level, in other words, quasi-continuous distribution.
If trap centre is considered to be formed by electron injection only, current formed by detrap
electrons in the external circuit is:
0
0
1,
00
12 , d
0
, d d
, d2
T
nc T
v
T
nc T
v
e E T dTE l
t nE
e E T TE
t nE
xJ T ef E N E e E T e x E
d
elf E N E e E T e E
d
(8)
CAPITOLO 5
37
Where: l——injection depth of electron/m; f0(E)——constant probability function of ebergy
level E to be occupied by electron; Nt(E)——trap energy level density function/(eV·m3)-1 ;
d ——sample thickness/m; e ——unit electronic charge/C; T0 ——initial temperature/K; T
——test temperature/K; β ——heating rate/K.s-1; E ——trap energy level/eV; Ev ——
valence band energy level/eV; Ec ——conduction band energy level/eV.
At T temperature, probability of electrons excited from trap energy level E to conduction
band en(E,T) is
, 1E
kTne E T e
(9)
Where: τ ——relaxation time/s; ν ——frequency factor of trap electron escaping/s-1, usually
ranges from 1012 to 1014 s-1; k ——Boltzmann constant.
Introduce a new function:
0
0
1, d
1d
1 , ,
T
nT
ET
kT
T
e E T T
n
E e TkT
G E T e E T e
e e
(10)
This function reflects the influence of trap energy level distribution to external circuit current
under temperature T. Using integral approximation, G1(E,T) can be simplified to G2(E,T):
2
2 ,
E
kTkTE e
EkTG E T e e
(11)
Similarly, approximation G1(E,T)、G2(E,T) by δ function, get:
, m mG E T A E E E (12)
A is the function of Em. Substitute equation (12) to equation (8):
0 2
2/m t m m
df E N E J T A E
el (13)
CAPITOLO 5
38
The function distribution of trap energy level can be obtained by TSDC test result once A(Em)
is solved. The trap density calculated by this method may be less than the true value, because
all the traps are assumped to be full fo electrons at the initial condition.
Figure 17 is the result. Average teap density can be obtain by integrating trap energy level
density by trap depth.
Figure 17: Trap Energy Level Distribution
5.3.2 Trap Energy Level Distribution
Here the trap energy level distribution is discussed separately.
For micron ZnO modified PI, results are shown in Table 7 and Figure 18.Shallow trap
energy level left shifted while deep trap energy level right shifted, which means that micron
ZnO modification decreases shallow trap energy level while increases deep trap energy level.
Besides, average shallow trap density, average deep trap density and average trap density
increased, which means that micron ZnO modification increases trap number. The energy
level density peak of 5wt%μm’s deep trap is relatively shorter and fatter, which means the
energy level distribution is more disperse. On the other hand,there is no obvious rules
between those trap parameters and mass fraction of ZnO.
CAPITOLO 5
39
(a) shallow trap (b) deep trap
Figure 18: Trap Level Distribution of μm ZnO/PI
Table 7: Trap Depth and Average Density of μm ZnO/PI
Sample Average trap
density/×1017
m-3
Average
shallow trap
depth/eV
Average shallow
trap
density/×1016
m-3
Average deep
trap depth/eV
Average deep
trap
density/×1017
m-3
PI 4.47 0.69 2.72 1.27 4.20
1wt% 5.09 0.67 2.73 1.35 4.80
2wt% 4.56 0.60 2.96 1.30 4.22
3wt% 5.24 0.66 3.64 1.30 4.88
5wt% 4.58 0.67 3.85 1.30 4.30
For nano ZnO modified PI, results are shown in Table 8 and Figure 19. Shallow trap energy
level depth of nano ZnO modified PI is around 0.7eV for different ZnO mass fraction
samples, while deep trap energy level depth increased. For average shallow trap density, all
modified samples except 1wt%nm had a larger value. For average deep trap density,
1wt%nm and 2wt%nm had larger values while 3wt%nm and 5wt%nm had smaller values.
The reason maybe that when the additive amount is huge, aggregation effect weakens the
interface trap effect. The properties and shape of nano particles is the determinant of its
interaction, dispersity and ability of heterogeneous nucleation with polymer, which is
relative to the interface trap effect. So micro ZnO modification differs from nano ZnO
modification. Other reseachers also got silimar result[19]. Peak value of deep trap energy
CAPITOLO 5
40
level density of all nano samples all decreased, which means nano modification makes deep
trap energy level more disperse. Among all the samples, 3wt%nm has the lowest value of
average trap density, which means it has the lowest number of traps in the same volume.
(a) shallow trap (b) deep trap
Figure 19: Trap level distribution of nm ZnO/PI
Table 8: Trap depth and average density of nm ZnO/PI
Sample Average trap
density/×1017
m-3
Average shallow
trap depth/eV
Average shallow
trap density
/×1016
m-3
Average deep trap
depth /eV
Average deep trap
density /×1017
m-3
PI 4.48 0.69 3.40 1.27 4.13
1wt% 4.63 0.69 3.35 1.30 4.36
2wt% 4.52 0.68 3.44 1.30 4.23
3wt% 3.60 0.70 3.52 1.30 3.87
5wt% 4.33 0.71 3.53 1.36 4.01
On the whole, modification brings in large amount of interfaces and deeper the deep trap
energy level, which makes it easier to catch space charge. Micron modification and low-
doping nano modification have larger average trap density than pure PI while high-doping
nano modification have smaller average trap density.
CAPITOLO 5
41
5.4 Brief Summary
Through TSDC test and the improved quasi-continuous distribution method, trap energy
level distribution and average trap energy level depth and density were obtained. The result
shows that deep trap energy level became deeper after modification because interfaces
between ZnO and PI bolcks chain movement. On the whole, micron modification and low-
doping (1wt%,2wt%) nano modification have larger average trap density than pure PI while
high-doping (3wt%,5wt%)nano modification have smaller average trap density.
CAPITOLO 6
42
6 Research on Conductance-Temperature
Characteristic of Modified Polyimide
There have been plenty researches about the nonlinear conductivity of modified polyimide.
In this chapter, temperature characteristics are studied, and the influence of particle size and
mass fraction of the additive is analyzed.
6.1 Test Platform of Nonlinear-Conductance Expriment
Test platform is shown in Figure 20.
Figure 20: Conducting Test Platform
Since the surface of spacecraft is usually in negative potential, here we use negative voltage
source. It can generate dc voltage from 0 to 60kV, and the resolution is 0.1 kV. The three-
electrode system is put inside an oven so that the conductivity test can be done under
different temperatures. Here we use a high resistance meter to measure the current folwing
through the sample. The 100MΩ and 50MΩ resistances are in series in the circuit in case
there is large current.
Before the test, samples were polished to 0.5mm thick, and wiped by alcohol. After that,
samples were put inside the oven at 120°C for 2 hours to eliminate residual charges. During
the test, voltage were generated from 1kV to around 21kV step by step, and each step was 2
kV. Each voltage step was kept for 1 min to charge the sample. After 1min, current was
CAPITOLO 6
43
recorded. Using formula (13) to calculate current density, and draw current density versus
electric field strength curves.
2 2
1 1
4
(D g) / 4 (D g)
V V V
V
I I IJ
S
(14)
Where: JV——current density/A/m2; IV——measured current/A; S——Effective contact
area/m2; D1——diameter of electrode/24 mm; g——gap between measuring pole and
shielding pole/2 mm.
6.2 Influence of Concentration of ZnO
First, samples were tested under room temperature. Using the curves of Current density j
versus electric field strength E to present samples’ conductivity properities. Figure 21 is the
results.
Figure 21: Conductivity Test Result at 25°C
The current density of pure sample is linearly increased as field increased, and the value is
around 10-8A/m2 order of magnitude. Dielectric’s conductivity characteristics under strong
electric field is that when field strength is near breakdown strength, trapped charged will be
pulled out of traps so that current inside dielectric will increase rapidly. In this case, the
material transients from linear area to nonlinear area. This procedure is irreversible. In the
test, the pure sample remains in linear zrea, which means the electric field is not strong
enough to let the transition happen. From previous test results, dc breakdown field strength
of pure polyimide is around 170kV/mm, which is much larger than the test field. On the
other hand, modified polyimide samples have nonlinear conductivity. The current density
increases slowly and linearly as electric field increases, and when fiels strength reaches a
CAPITOLO 6
44
certain value, current density increases rapidly. The transient field is called threshold field.
After threshold field, current density is around 10-6A/m2, which is two orders of magnitude
larger than pure polyimide sample.
For samples modified by different mass fractions of micron ZnO, the values of current
density is similar before the threshold field. But after threshold field, current density of
samples shows differences and so as the threshold field. The sample modified by 3wt% μm
ZnO has the lowest value of threshold field, which is 33 kV/mm, but its current density after
threshold field is small and the curve is not so steep with respect to other samples, which is
around 10-7A/m2. Combine the experimental results Table 7 of the previous chapter,
3wt%μm sample has the largest average trap density, which means that under the same
volume its has the largest number of traps. With large quantity of traps, it can capture large
quantity of carriers and reduce the carrier mobility, thus the current density is reduced.
For samples modified by different mass fractions of nano ZnO, the values of current density
show differences before the threshold field. The sample modified by 5wt% nm ZnO has the
lowest value of threshold field, which is 24 kV/mm. However, the current density of the
sample modified by 3wt% nm ZnO after threshold field is twice larger than that of 5wt%
nm, and is the largest among all the samples, with threshold field of 30 kV/mm. In addition,
the steepness of its curve is moderate, not as steep as 1wt%nm and 2wt%nm. In conclusion,
3wt%nm ZnO/PI has the best nonlinear conductivity behavior. Similarily, combine the
experimental results Table 8 of the previous chapter, 3wt%nm sample has the lowest
average deep trap density and average deep trap energy level, which means that it can capture
the smallest number of carriers and the detrapping field is the lowest among all the samples.
The nonlinear conductivity behavior is well corresponding to the result of trap experiment.
Threshold of each sample is in Table 9.
Table 9: Threshold Field of Each Sample at 25°C
Sample
μm ZnO/PI nm ZnO/PI
1wt% 2wt% 3wt% 5wt % 1wt% 2wt% 3wt% 5wt %
Threshold field
/kV/mm 47 47 33 41 45 36 30 24
CAPITOLO 6
45
6.3 Influence of Grain Size of ZnO
In order to compare the nonlinear conductivity characteristics of micron modified polyimide
and nano modified polyimide, current density curves of micron and nano modified samples
at the same ZnO addition mass fraction were draw in the same figure.
Figure 22: Nonlinearity Comparison of μm and nm ZnO/PI (25°C)
We can see that nm ZnO/PI has better nonlinear performance than μm ZnO/PI. It has lower
threshold value and larger current density after threshold. For low doping samples(1wt%,
2wt%), current density before threshold and curve steepness after threshold are silimar of
micron and namo modification, while for high doping samples (3wt%, 5wt%), current
density of nano modified samples before threshold are larger than that of micron modified
samples, and the curve steepness are quite different. In general, 3wt%nm and 5wt%nm have
relatively better nonlinear conductivity properity.
6.4 Influence of Temperature
3wt%nm ZnO/PI and 5wt%nm ZnO/PI samples were used to do the temperature test. Testing
temperatures are: 40°C,60°C,80°C ,100°C The results are as following.
CAPITOLO 6
46
Figure 23: Conductivity Test Result at Different Temperatures
Table 10 Threshold Fiels at Different Temperatures
Iteam 3wt%nm 5wt%nm
Temperature/°C 40 60 80 100 40 60 80 100
Threshold/ kV/mm 25 - - - 25 25 25 25
From Figure 23 we can see that as temperature increases, current density increases obviously
even before threshold, which means thermal hopping conduction takes effect. Higher
temperature increases carrier energy, so that carrier mobility increases, thus current density
increases.
When temperature is at 60°C and above, the boundary of ohm area and nonlinear area of
3wt%nm is not so obvious. It is not easy to read the threshold since current density increases
nonlinearly at low field.
For 5wt%nm, it has a clear transient from linear to nonlinear at different temperatures, and
the threshold value is around 25 kV/mm, which is similar to the value at room temperature.
It means that temperature doesnot change the threshold of 5wt%nm. On the other hand,
current density before threshold increases remarkably as temperature increases.
6.5 Research on Nonlinear Conductivity Mechanism
The relationship between current density and electric field strength is shown inFigure 24.
The dash line represents j-E curve of normal dielectric material, while the solid line
represents j-E curve of modified dielectric material which has nonlinear conductance.
CAPITOLO 6
47
Figure 24: Nonlinear Conductivity Curve
The nonlinear conductivity zone is before breakdown field. Materials with nonlinear
conductivity can have relatively large current density under low electric field. It may have
different conductivity mechanism. The following is the analysis of the nonlinear
conductance mechanism at different temperatures.
6.5.1 Mechanism of Nonlinear Conductivity at 25°C
For polymer, ions are the mainly conductive carriers under low field. For pure polyimide,
ions are derived from impurities, so conductance is low because pure polyimide has a little
impurities. For modified polyimide, ions are also derived from the ionization of ZnO, so
current density of modified polyimide is larger than that of pure polyimide at low field.
Current density of ion conductivity depends only on carrier concentration and mobility, so
current density and electric field strength obey the Ohm's law. Under high field, conductive
carriers are mainly the electrons. Potential barrier between two small crystalline region[6]
and contact-potential barrier between the modifier and base material will tilt under the
influence of electric field. In this condition, electrons can easily jump over potential barriers
and form the electron hopping conduction. Electrons in the modifier first jump over the
contact-potential barrier into the polymer, then jump over potential barriers between
amorphous regions when migrating inside the polymer. At last, electrons jump over the
contact-potential barrier from polymer to modifier when they reach another interface. This
process is repeated to achieve the conducting of modified polyimer. The repeated
unit ”modifier-interface-polymer” forms the path of hopping conduction[12].
The conducting mechanism of polymer is complicated and yet has not formed a complete
theoretical system. For modified polymer, the mechanism is even more complex and diverse.
CAPITOLO 6
48
In this work, mathematical method was used to do data processing, and test result was
compared to several classic conductivity theories to make the fianl conclusion.
1) Schottky effect or Pool-Frankel effect
For Schottky effect or Pool-Frankel effect, the logarithm of current density lnj is linear to
E . Polt lnj vs E , shown in Figure 25.
Figure 25: Schottky Effect Linear Fit at 25°C
The linearity of lnj- E is not good, so the mechanism is not or is not only Schottky effect or
Pool-Frankel effect. Consider that ZnO introduces the interfaces of ZnO and PI, which is
highly conductive[17], so the mechanism is more complecated than pure polymer
conduction.
2) Tunneling effect
For Tunneling effect, ln(j/E2) is linear to(1/E). If electric field does not change much, E2 can
be neglected. Thus, ln(j) is approximately linear to (1/E). Polt lnj vs1/E, shown in Figure 26.
Figure 26: Tunneling Effect Linear Fit at 25°C
CAPITOLO 6
49
The linearity of lnj- 1/E is quite good, so Tunneling effect may be the mechanism. Compared
with micron and nano ZnO modified polyimide, it is found that the nonlinear conductance
of micron ZnO modified polyimide is independent of the amount of modification, while for
nano ZnO modified polyimide, the curves of different modifications are different.
3) Space charge limited current (SCLC) effect
In SCLC, the logarithm of current densit lnj is piecewise linear to the logarithm of electric
field ln E . Slope in the ohm zone is 1 while in the trap zone is 2.
Peform linear fit to lnj - ln E of each sample. It is found that curves have three linear zone
with different slopes. Calculate slopes of each linear zone, shown in Figure 27.
CAPITOLO 6
50
Figure 27: SCLC Linear Fit at 25°C
For the first segment of linear section, slopes are approximately equal to 1 except 1wt%nm
and 5wt%nm. Slopes of these two samples are less than one, which may caused by
experimental error. In general, the first linear section meets the ohm zone of SCLC. At this
state, part of the charges flowing into the material participate in conducting, while others are
trapped. Macroscopically, it shows ohmic characteristics.
For the second segment of linear section, slopes are approximately equal to 2 except
2wt%μm, 5wt%μm and 5wt%nm. In general, it meets the space charge limited current zone
of SCLC effect. At this state, traps are all filled with charges. Macroscopically, injected
current equals to the current flowing out, and space charge limited current begins to work.
For the third segment of linear section, slopes are all larger than 3. It meets the step from b
to c in Figure 4. At this state, charges are detrapping from traps and participating in
conducting, thus current density increases rapidly. At this time, the slope reflects the flatness
of the nonlinear conductance. The higher the slope, the larger the steepness, the more
difficult to control the discharge. It can be seen that solpes of 3wt%μm, 3wt%nm and
CAPITOLO 6
51
5wt%nm are around 5 while others’ are above 10. It means that those three samples have
better nonlinear flatness, so diacharge can occur in a relatively mild way. The phenomenon
that slope after threshold is greater than 2 is also found in others' experimental studies[21].
Area c in Figure 4 , which is the state when all charges have detrapped so that slope stabilised
at 2, does not show up. The reason maybe that electric field are not high enough.
In general, conduction mechanism of modified polyimide fits SCLC effect. The tramsform
field is in Table 11.
Table 11: Transient Field of SCLC at 25°C
Sample
μm ZnO/PI nm ZnO/PI
1wt% 2wt% 3wt% 5wt % 1wt% 2wt% 3wt% 5wt%
Eab/ kV/mm 22.20 20.08 18.92 18.36 20.08 20.08 24.05 15.80
ET/ kV/mm 47.94 47.00 41.68 40.45 44.25 36.60 30.57 29.08
Compared to Table 9, ET is colsed to the threshold field of nonlinear conductivity. It means
that nonlinear conductivity is caused by charges detrapping from traps and participating in
conducting. Due to the modifier ZnO, a large number of interfaces are introduced, resulting
in an increase in the number of traps in the sample. When electric field strength is low, part
of the injected charges are trapped and donnot participate in conducting. At this time,the role
of these traps can been as "charge storage". When electric field arrives a certain value(ET),
charges will gain more energy, and interface barrier will tilt, which makes trapped charges
detrapping and current density increasing. It also explains the reason why pure polyimide
samples do not have nonlinear conductivity: the "charge storage" ability is weak due to that
traps introduced by interfaces are very few, so the number of trapped charges which can
participate in conducting is not enough.
According to the result of TSDC in chapter 3, high-doping nano modification have smaller
average trap density, which is contrdict to the analysis above. In this case, the properties of
nano particles should be considered. At the same doping concentration, the number of nano
particles is far greater than the number of micron particles because of small particle size. As
a result, polymer layer between two nano particles is so thin that partial strong field is easily
generated, giving rise to the tunnel effect. Paper[13] also validates this theory. So the
CAPITOLO 6
52
conducting mechanism of nano modified polyimide should be explained by both SCLC and
tunnel effect.
6.5.2 Mechanism of Nonlinear Conductivity at Different Temperatures
Likewise, experimental data were processed against basic conductive theories.
1) Schottky effect or Pool-Frankel effect
Figure 28 is the lnj- E curves of 3wt%nm and 5wt%nm at different temperatures. From the
figures we can see that the linearity is not good except the sample of 3wt%nm under 80°C
and 100°C.
Figure 28: Schottky Effect Linear Fit at Different Temperatures
2) Tunneling effect
Figure 29 is the lnj-1/E curves of 3wt%nm and 5wt%nm at different temperatures. Curves
under strong field is consistent with the tunnel effect. Considering that tunneling effect is
independent of temperature[6], the curves are up and down with the change of temperature,
so the tunneling effect can not be explained at this time.
CAPITOLO 6
53
Figure 29: Tunneling Effect Linear Fit at Different Temperatures
3) SCLC effect
Figure 30 is the lnj- lnE curves of 3wt%nm and 5wt%nm at different temperatures. Similarly,
piecewise linear fitting was performed and slopes of each linear zone were calculated.
CAPITOLO 6
54
Figure 30: SCLC Linear Fit at Different Temperatures
For 3wt%nm, at 40°C, it has three linear sections with different slopes. At other temperatures,
it only has two linear sections with slopes of 1 and 2. Looking back Figure 4, when
temperature increases, traps become shallower, and line b comes closer to the dash line. Line
c becomes shorter and shorter, until line b overlaps with the dash line. At this time, traps are
all “disappeared”, or the barrier height becomes zreo. Combining with experimental results,
it means that when tempearture is 60°C and above, conducting mechanism is SCLC effect
with no internal traps, i.e. Calder’s Law. In Figure 30, slopes of 3wt%nm at 60°C and above
before Eab is larger than 1. It means that current density changes as electric field changing
before Eab, which explains the phenomenon that ohm zone becomes shorter and shorter. At
high temperatures, slopes of current surge area is no more than 2, which means good
nonlinear flatness.
For 5wt%nm, it has three linear zone at each temperature. From TSDC test result in Table
8, the average trap density of 5wt%nm is 4.33×1017m-3, much larger than 3.60×1017m-3 of
3wt%nm. And average deep trap energy level of 5wt%nm is 1.36 eV, deeper than 1.30 eV
of 3wt%nm. So the influence of temperature to 5wt%nm is not so strong as to 3wt%nm
CAPITOLO 6
55
because 5wt%nm has more dopping amount. And slopes of the third linear section is between
5-7, which means quite good nonlinear flatness.
Transient field is in Table 12.
Table 12: Transient Field of SCLC at Different Temperatures
Iteam 3wt%nm 5wt%nm
Tempweature/ °C 40 60 80 100 40 60 80 100
Eab/ kV/mm 17.81 17.64 27.38 21.12 14.88 14.58 14.88 14.58
ET/ kV/mm 30.57 - - - 25.03 27.38 24.78 24.78
6.6 Brief Summary
In this chapter, conducting tests under different temperatures were carried out, and
conducting mechanism was discussed.
1) A conducting test platform which can be used at different temperatures was built.
2) Under room temperature, modified samples has nonlinear conductivity. Among all the
samples, nano modified polyimide has better nonlinear conductivity with lower threshold
field. The nonlinear conductivity of 3wt% nm is the best, with the threshold field being
30kV/ mm. Then several classic conducting mechanism were compared to test result, and it
is found that the nonlinear conductivity behavior is in consistant with the SCLC effect. The
occurrence of the nonlinear conductivity is due to the detrapping of space charges. At last,
the conductivity mechanism of nano and micron modified polyimide is discussed in detail
combining tunnel effect and trap distribution.
3) Conducting tests of 3wt%nm and 5wt%nm at 40°C, 60°C, 80°C, 100°C were done.
Current density increases obviously as temperature increases. It is assumed that carrier
energy is increased and trap barrier is decreased due to temperature, so trapped charges are
more likely to leap through the barrier or excite from the trap and participate in the
conduction. The sample of 3wt%nm does not have an obvious ohm zone at 60°C and above,
and it is considered that it obeys SCLC without traps inside. While 5wt%nm shows nonlinear
conductivity characteristics at all temperatures, and the threshold field is around 25kV/mm.
CAPITOLO 6
56
It is considered that it obeys SCLC with traps inside. Additionaly, the influence of
temperature to trap energy level of 5wt%nm is less siginificant than that of 3wt%nm.
CONCLUSIONI
57
7 Conclusion and Perspectives
7.1 Conclusions
In this work, polyimide was modified by micron and nano sized ZnO respectively, and the
ZnO content was 1wt%, 2wt%, 3wt%, 5wt%. Dielectric properties, trap energy level
distribution and conductivity in different temperatures and electric fields were measured for
each samples. Finally, the following conclusions are obtained:
1) The addition of ZnO reduces volumn resistivity of mofidied polyimide, and low dopping
(1 wt%, 2 wt%) makes relative permittivity and dielectric loss lower while high doping (3
wt%, 5 wt%) makes the relative permittivity and dielectric loss higher. DC breakdown
strength decreases after modification, but still meet the insulation requirements. Among all
the samples, the change of 5 wt% nm modified sample is the largest, and it is considered that
nano particles are agglomerated to form "nano bridge" due to the large doping amount.
2) TSDC experiment shows that modification deeper the trap depth of polyimide, the reason
may be the introduction of more defects in the modifier. Micron modification and low
dopping (1 wt%, 2 wt%) nano modification increased trap density of the modified sample,
while high doping (3 wt%, 5 wt%) nano modification reduced trap density.
3) At room temperature, modified samples have obvious nonlinear conductivity
characteristics, and nano modified samples have better nonlinear effect than micron samples.
3wt% nm has the best nonlinear conductivity performance, with threshold field of 30kV/mm.
It is found that the nonlinear conductivity characteristics of each sample are in good
agreement with the space charge limited current theory by piecewise fitting of current
density and electric field. The nonlinear conductivity is caused by the space charges
detrapping from traps and involving in conducting. For nano midified samples, tunneling
effect also contributes in the conducting mechanism.
4) As temperature increases, current density increases significantly. The reason is that
temperature increases carrier energy and also decreases trap energy level, so space charges
are more likely to jump over barriers or excite from traps to participate in conducting. Ohm
zone of 3wt% nm becomes shorter at 60°C and above, which meets the sclc theory of the no
trap condition. The threshold field of 5wt% nm does not change with temperature, which is
CONCLUSIONI
58
about 25 kV/mm, and the conductivity characteristic meets the sclc theory with traps. It is
believed that when the doping amount is large (5wt%), the effect of temperature on the trap
energy level is not as significant as 3wt%nm.
7.2 Outlooks
Further research can be carried out from the following aspects:
1) Modification process and the correspongding performance tests can be repeated to
improve the modification technology.
2) Space charge experiment can be carried out to inspect the positive and negative polarity
of space charges and their distribution position. Internal charges can be better analysised
combining with TSDC experiment.
3) Conducting tests at lower temperatures and be carried out to have a more comprehensive
study on temperature characteristics of nonlinear conducvitity.
4) In order to simulate the space environment better, tests can be done under repeated
irradiation environment and thermal-cold cycling conditions.
BIBLIOGRAFIA
59
Bibliografia
[1]Qin Xiaogang, Zheng Xiaoquan, Wang Li, Life Evaluation Technology of Spacecraft
Dielectric on Electrical Parameters, Vacuumand Cryogenics, 2007, 13(4), 198-201.
[2] Quan Ronghui, Zhang Zhenlong, Han Jianwei, Huang Jianguo, Yan
Xiaojuan,Phenomenon of Deep Charging in Polymer under Electron Beam Irradiation,
Acta Physica Sinica, 2009, 58(2), 1205-1211.
[3] Bie Chengliang, Research on Dielectric Property of FR4 Circuit Board after High
Energy Electron Radiation and Non-linear Conductance Modified, Master Thesis,
Xi’an: Xi’an Jiaotong University, 2009.
[4] Zeng X. C., Bergman D. J., Hui P. M., et al. Effective-medium Theory for Weakly
Nonlinear Composites. Physical Review B Condensed Matter,1988,38(15):10970-
10973.
[5] Wu Jiang, Bai Jingjing, Shen Bin, Zheng Xiaoquan, Formation Mechanism of Anti-
deep-charged Modification for Space Dielectric, Chinese Space Science and
Technology, 2010,1(02):49-54.
[6]Jin Weifang, Dielectric Physics, Xi’an Jiaotong University Press, 1997:114-115,122.
[7] Zhong Lisheng, Li Shengtao, Xu Chuanxiang, Engineering Dielectric Physics and
Dielectric Phenomenon, Xi’an Jiaotong University Press, 2013:156.
[8] Zhang Zhenjun, DC Surface Flashover Characteristics of Polyimide in Simulated Space
Environment, Master Thesis, Xi’an: Xi’an Jiaotong University, 2014.
[9] Kang Yali, Research on Deep Charging Property of Typical Space Materials, Master
Thesis, Xi’an: Xi’an Jiaotong University, 2011.
[10] Donnelly K. P., Varlow B. R. Non-linear dc and ac Conductivity in Electrically
Insulating Composites. IEEE Transactions on Dielectrics and Electrical Insulation,
2003,10(4): 610-614.
[11] Guo Wenmin, Han Baozhong, Li Zhonghua, An Experimental Investigation on the
Influence Factors of The Electric Field-Dependant Conductivity of LDPE/ZnO
Composites, Journal of Functional Materials, 2009, 40(06): 943-945.
[12]Guo wenmin, Research on Non-linear Conductive Characteristics and Mechanisms of
Polyethlene Composites Filled with Inorganic Filler, Doctor Thesis, Harbin: Harbin
University of Science and Technology, 2010.
[13] Liu Chenyang, Li Kangning, Zheng Xiaoqian. Discussion on Non-Linear Conductivity
Characteristics With Space Charge Behavior of Modified Epoxy for Spacecraft. IEEE
Transactions on Nuclear Science,2016,63(52):2724-2730.
[14] Donzel L., Greuter F., Christen T. Nonlinear Resistive Electric Field Grading Part 2:
BIBLIOGRAFIA
60
Materials and Applications. IEEE Electrical Insulation Magazine,2011,27(2):18-29.
[15]Lei Weiqun, Wu Jiang, Peng Ping,et al.Conduction mechanism analysis of modified on
polyimide composite. Journal of Beijing University of Aeronautics and Astronautics,
2015,41(6): 1049-1054.
[16] Wu Guoguang, Polyimide Film and It's Applications in Aerospace, Information
Recording Materials, 2012, 13(01): 28-34
[17]Li Shengtao, Yin Guilai, Wang Weiwang, Li Jianying, Investigation Development and
Consideration on Nanodielectrics, Proceedings of the 13th National Conference on
Engineering Dielectric, Shaanxi, Xi’an. 2011
[18] Wang Hui, Research on Dielectric Properties of Polyolefin Composites, Master
Thesis, Harbin: Harbin University of Science and Technology, 2014.
[19]Tian Fuqiang, Investigation on the Trap Characteristics and Electrical Properties of
Polyethylene Based Nanocomposite, Doctor Thesis, Beijing: Beijing Jiaotong
University, 2012.
[20] Bucci C., Fieschi R. Ionic Thermoconductivity. Method for the Investigation of
Polarization in Insulators. Physical Review Letters,1964,12(1):16-19.
[21]Lan Li, The Effect Of Particle Surface Modification On Dielectric Properties In Polymer
Nanocomposite, Master Thesis, Shanghai: Shanghai Jiaotong University, 2012.
AUTHOR’S PUBLICATION
61
Author’s Publication
AUTHOR’S PUBLICATION
62
AUTHOR’S PUBLICATION
63
AUTHOR’S PUBLICATION
64
AUTHOR’S PUBLICATION
65
AUTHOR’S PUBLICATION
66
AUTHOR’S PUBLICATION
67