Polymer under AC Voltage

3 downloads 0 Views 4MB Size Report
Kai Yang, Guan-Jun Zhang, Wen-Bin Zhao, Zhang Yan. State Key Laboratory ofElectrical Insulation and Power Equipment, School ofElectrical Engineering,.
XXIInd Int. Symp. on Discharges and Electrical Insulation in Vacuum-Matsue-2006

Surface Electroluminescence Phenomena from Polymer under AC Voltage Kai Yang, Guan-Jun Zhang, Wen-Bin Zhao, Zhang Yan State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi 710049, China to the theory of the isothermal relaxation current and the measurement ofthe surface potential.

Abstract- Electroluminescence (EL) Phenomena from polymers has been investigated under 50Hz ac electrical stress. Because the light emission is very week, so we employ a sensitive Photon Counting Module (PCM) and put the sample into vacuum chamber to avoid the influence of gases or impurities absorbed on the surface of the sample. We use three kinds of materials such as PTFE, LDPE, and PP in this experiment. Their surface trap parameters are measured according to the theory of isothermal relaxation current (IRC). By that method we acquire the trap energy level and trap density of both electrons and holes distributing in the surface layer of each insulating polymer. We find that there is some relationship between the surface trap parameters of solid polymer insulating materials and their molecule structure, and surface trap has influence on light emission from polymer insulating material.

II. EXPERIMENTAL The dimension of samples of PTFE, LDPE, PP is lOx5xO.5 mm3, before test it is put into absolute ethyl ethanol and washed by ultrasonic cleaner equipment about 15 minutes. The stainless electrodes are used and butted to the surface of the samples 10mm away. The samples are tested individually inside a light-tight test chamber, which is evacuated tolO-4 Pa by a mechanism pump and a turbo molecule pump, shown in Fig. 1.

I. INTRODUCTION

Electrical breakdown of polymeric insulation is a prime concern of the electrical generation, transmission and electric cables. Polymer based insulation is becoming more popular, particularly for high voltage cables, owing to their excellent physical and electrical properties. Polymers insulating materials are now widely used in cable insulation for its relative cheapness and ease of processing. Electrical breakdown in polymeric materials is often the result of slow material degradation in regions of high electrical stress that eventually gives rise to the formation of the voids [1]. EL technique is sensitive enough to reveal the very early stages of insulation degradation, prior to voids inception and partial discharges [2-4]. A photon counting module SPCM-AQR, which is a self-contained module that detects single photon of light over the wavelength range from 400 nm to 1060 nm, is employed to detect EL emitted from the surface of polymers, it requires a +5 volt power supply. A TTL pulse, 5 volts high in a 50Q load and 30 ns wide, is output at the rear BNC connector as each photon is detected. Many researchers have studied EL under 50Hz ac electrical stress [2,3], and models for EL based on bipolar charge recombination have been proposed [4]. These studies have demonstrated that EL is results of charge injection and recombination, which is related with the surface trapping centers. In this paper we design a device to acquire the trap parameters according

1-4244-0191-7/06/$20.00 C 2006 IEEE.

-

Fig. 1 Experimental system structure for electroluminescence measurement

Figure.2 Experimental setup used for dc corona and surface potential measurement

The trap parameter is acquired by a setup

833

-

that we designed according to the theory of IRC and measurement of surface potential as shown in Fig.2. In this study, the sample is stuck firmly to a stainless steel plate electrode with conductive silica grease. The needle-plane corona discharge is employed to charge the samples. The upper side of each sample is subjected to a dc corona discharge in atmosphere, and the other side is grounded. The applied voltage is +/-4kV and the charging time is 5 min. After charging, all needle electrodes are taken away, and the two sides of the sample are short-circuited for a short time to remove the free charges drifted to the sample surface. Then quickly put the vibration capacitive probe above the sample surface 2mm away. Thus the surface potential can be measured and recorded in computer.

time, assume that p is nearly 0. Because of 2d >>d, then

Vd=p£r

From Eq. (4), it is concluded that the surface potential is dominated by the space charges near to the surface layer not contacted to the ground. Obviously the free charges on the surface are also contributed to the surface potential. Thus, in order to verify the surface trapping parameters correctly, these free charges should be removed by the shorting of two sides of each charged sample. Under a constant temperature, the charge carriers trapped in trapping centers with low energy are detrapped more easily, and then the carriers trapped in deep trapping centers. The isothermal relaxation current reflected the distribution of traps with different energy levels. According to the theory analysis given by Simmons [5,6,7], the relationship among the level and density of trapping centers and the IRC density are as equation (5):

III. RESULTS & DISCUSSION After each sample is charged for 5 min, it is supposed that space charges inside the sample would distribute as shown in Fig.3, Theoretical analysis and experimental results proved that the space charges due to charge injection into an insulating material mainly existed in its surface layer of 1-2Vtm [8,9,10]. Simplify the space charges in Fig.3 as a one-dimension distribution along the normal direction of a sample. The surface potential Vs predominated by the trapped charges can be written by (1):

PA

P+c +

Et = kTln(Q) = qd 2t fo (Et )Nt (Et)

--

-

-

-d- 36

+ + + + + +

- Negative

++ +

+ +

J(t) -0

Fig. 3 The space charge distribution trapped inside an insulating material in charging equilibrium 1

0 £O£r

°d xp(x) d x(1

(1

Assume the charge density in the region of 0-6 and d-&-d as p+ and p, and in the region of 3-6-d as p, respectively, then

VS

= O r

[ i°

p+dx+f

xpdx+fd_ 5xp

dx]

(2)

According to the Netwon-Leibnitz equation

VS

x2pP

+

x2pd

+-x2P_

=oer d V, (t) d dt

(6)

According to Eq. (6), and based on the measurement of surface potential, the current density can be calculated numerically. Then the trapping parameters can be derived from Eq. (5), the polynomial Fit curves is shown in Fig.4. From the Fig.4, we can see that holes and electrons traps on the surface LDPE are much similar, and it is different from the other polymers. The peak value of electrons trap of pp is much higher than that of holes trap, and the depth of electrons trap of PTFE is deep than that of holes trap. We think that there is some relation with the structure of polymer. On the molecular of LDPE, the main chain is carbon atoms, where 2 hydrogen atoms surround, the structure is symmetry, which would maybe determinate the traps density of holes and electrons. The structure of PP is difference from PE that one of hydrogen atom branch is replaced by a methyl, which induces the asymmetry of PP structure. Because of that reason, the curve shape of PP is different from LDPE. We can see that the distribution of electrons density Vs

space charge +Positive space charge

Is=

(5)

Where, Et, Nt(Et) andfo(Et) means the level, density and initial occupancy of trapping centers, respectively, k is the Boltzmann constant, q is the electron charge, yis the vibrating frequency of electron, and T, t and d indicates the temperature, time and thickness of sample, respectively. Let us investigate the relation between the decay of surface potential and the IRC I(t) =CdVs(t)/dt. The current density can be written as equation (6):

-d

--

(4)

£o

(3)

While using a negative needle-plane discharge as shown in Fig. 2, it is believed that p >>p+. At the same

834 -

energy differs with that of holes. To the PTFE, one of hydrogen atom branch is replaced by a electronegative fluorine atom, which make the whole characteristics of PTFE electronegative, so the energy of electron is deeper than that of holes and the density of electrons is should be more than that of holes, but the curve it is not obvious, we think it is because of the device limit that we cannot measure the deep energy density of electrons and holes. We just make the assumptions of that and should do more work to explain the phenomena. ,40 -

C).

C.) ; C)

6000

_

4000

Initial Voltage: 4.5kV

2000

0

(a)

-Hole - Electron

In

(a)

8000

---

1-111-11,11".."I'll'.1-

0

200

400

600

800

1000

1200

1400

Time(s)

ID

(b)

10000

c-z 30to

20

8000

-

0

-

10

-

0*

.5

20

ct

6000

0.7 8

0.80

0.82

0.84

0.86

0.88

0.90

0.92

0.94

4000

0.96

Energy level of trapping centers, Et (eV)

-

Initial Voltage: 4.0kV

2000 20 18

-

-

(b)

Hole Electron

l1

u

200

400

600

16

>14

800

1000

1200

140C

Time(s) -

1 0000

.12

(C)

a1)

10

X 8

-

4

-

2

-

8000 C).

C.)

6000

;

I~~~~ 0 4000 0.84 0.86 0.88 0.90 0.92 0.94 0.96

15i

Energy level of trapping centers, Et (eV)

Initial Voltage: 4. 1kV

2000 -

12 -

-

C-)

(c)

Hole Electron

0 0

10-

400

600

800

1000

1200

1400

Time(s) Fig.5The EL from polymers (a) LDPE, (b) PP, (c) PTFE

8-

.

200

ct

C)

o 610000

.

-PP

-

-PTFE -LDPE

49000 -

20

*

0.80

0.82

0.84

0.86

0.88

0.90

0.92

0.94

Energy level of trapping centers, Et (eV)

0.96

0.98

C.) v

7000.-

5000-

If we integrate Fig.4 with the range of energy, we can get the surface trap density of electrons and holes trap, as listed in Table 1.

4000

-

3000

-

4.0

4.2

4.4

4.6

4.8

5.0

5.2

5.4

5.6

5.8

Voltage(kv)

Table 1 surface trap density of electron and hole trap Electron trap density Hole trap density Polymer ((Net) (x1018 m3)) 1.3411 0.3441 0.6797

8000-

.~;6000

Fig.4 The density of trapping centers to energy level, (a) PP, (b) LDPE, (c) PTFE

LDPE PP PTFE

C-

Fig.6 The EL scatter

((Nht) (XIOI'm_3))

curves

polymers Vs voltage

Fig.5 shows the EL curves of LDPE, PP. PTFE, the initial voltage of them is individual 4.5kV, 4.0kV, 4.1kV, in order to compare their

1.2522 0.3395 0.5768

-

835

-

characteristic we separately get the Fig.6 and Table 1 from Fig.5 and Fig.4. Comparing the EL from PTFE to LDPE or PP to LDPE, we find that the average EL from LDPE is larger than that from PTFE and PP, but in table 1 we find the density of them is inverse ratio to EL. So we think that it is must be some relationship with the charge movement in the surface of polymers and trap parameters. When the electrical stress is enough, the charge will move in the polymers, some of them will be capture by the traps in polymer, the deep traps will be firstly filled in, then the lower traps. So the trap density of polymers that is larger will be filled with more charge in deep trap but little in low trap. Under Ac voltage charge is injected alternately into polymer, light emission is recombination of electrons and holes. And the recombination process is inverse to the charge injection, it occurs firstly in low trap, so the polymer with small trap density will emit more light because that they have larger trapped charge in low traps than others. In this paper, the trap density of LDPE is largest so the average EL is smaller than PTFE and PP. But there is some difference between PTFE and PP from 4.0kV to 4.6kV, in this period we didn't find the similar phenomena, the reason is not clear and further study should be done to clarify it.

density polyethylene using acharge coupled device camera," IEEE. Trans. Dielectr. Insul. Vol. 13, pp. 168-181, 2006. [2] G. J. Zhang, Z.Yan, Yasuoka. K, et a]. "Electroluminescence from alumina ceramics surface under low AC electric field in vacuum," Proceeding of The 6th International Conference on Properties of Dielectric Materials, pp. 407 - 410, 2000. [3] K. Kojima, Y. Takai, and M. leda, "Electroluminescence in polyethylene terephthalate- II -ac voltage," J. Appl. Phys., Vol.2 1, pp.1436-1438, 1983. [4] J. M. Alison, J. V. Champion, and S. J. Dodd, "Dynamic bipolar charge recombination model for electroluminescence in polymer based insulation during electrical tree initiation," J. Phys. D: Appl. Phys., Vol. 28, pp. 1693_1701, 1995. [5] J. G. Simmons, M. C. Tam, "Theory of isothermal currents and the direct determination of trap parameters in semiconductors and insulators containing arbitrary trap distribution," Physical. Review .B, Vol. 7, Number. 8, pp.3706-3712, 1973. [6] J.G.Simmons, G.W.Taylor, "High-Field isothermal currents and thermally stimulated currents in insulators having discrete trapping levels", Physical .Revier .B, Vol.5, Number.4, pp. 1619-1629, 1972. [7] J.G.Simmons, G.W.Taylor, "Dielectric-Relaxation currents in insulators", Phys.Rev.B.5, PP.553-556, 1972 [8] P. A. Torpey, "Double-carrier injection and recom- bination in insulators, including diffusion effects", J. Appl. Phys., Vol. 56, No. 8, pp. 2284-2294, 1984. [9] S. R. Kurtz and R. A. Anderson, "Properties of the metal-polymer interface observed with space-charge mapping techniques", J. Appl. Phys., Vol. 60, No. 2, pp. 681-687, 1986. [10] G. J. Zhang, X. R Wang and Z. Yan, "Analysis of trapping parameters in surface layer of solid insulating materials", 2001 IEEE 7th International Conference on Solid Dielectrics, Eindhoven, the Netherlands, pp. 236-240, 2001.

IV. CONCLUSIONS In this paper, we design the measurement device of trap parameters according to the theory of isothermal relaxation currents and measure the trap parameters of several kind of polymers, we find that it has some relationship with the material structure .we also record the EL from polymers using a set up of Photon Counting System, comparing the trap density and the EL curve we believe that the trap distribution affects the EL characteristics of polymer, EL intensity is inverse ratio to the surface trap density. Trap density measurement will be improved in future and more detailed explain would be acquired in further study.

E-mail of KaiYang: yk200821IIA163.com E-mail of Guan-jun Zhang: gjzhangArmail.xjtu.edu.cn

ACKNOWLEDGMENT This work was supported by the New Century Excellent Talents in University (NCET-04-0943) and by the Excellent Young Teachers Program, Ministry of Education, China. The authors are grateful to Professor De-Min Tu at the State Key Laboratory of Electrical Insulation and Power Equipment, Xi'an Jiaotong University for his useful advice.

REFERENCES [1] S. J. Dodd, P. L. Lewin and K. I. Wong, " Phase resolved electroluminescence measurements in thin films of low

- 836 -