Experimental Study on Trap Characteristics of Nano-Montmorillonite Composite Pressboards

Qingguo Chen 1,2,*, Jiaxin Sun 1,2 ID , Minghe Chi 1,2,*, Jinfeng Zhang 1 ID and Peng Tan 2 1 Key Laboratory of Engineering Dielectrics and Its Application, Ministry of Education, Harbin University of Science and Technology, 52 Xuefu Road, Harbin 150080, China; jxsun2205@163.com (J.S.); zjinfeng1991@163.com (J.Z.) 2 The School of Electrical and Electronics Engineering, Harbin University of Science and Technology, 52 Xuefu Road, Harbin 150080, China; tp1728558632@163.com * Correspondence: qgchen@263.net (Q.C.); chiminghe1985@163.com (M.C.); Tel.: +86-451-8639-1601 (Q.C.); +86-451-8639-1625 (M.C.)


Introduction
The converter transformer is essential equipment for high voltage direct current transmission systems that transform alternating current (AC) electric energy into direct current (DC) electric energy and its dependability directly influences the stable operation of an energy system.As a key material for insulation, oil-paper insulation is widely used in converter transformers.Once this insulation loses efficacy, the transformer may malfunction irreversibly.Therefore, the insulating lifetime of oil-paper insulation approximately equals the practical functional lifetime of the transformer.Insulating structures withstand AC-superimposed DC voltage style and polarity reversal voltage, which concentrate the electrical field more than in a single electrical system [1][2][3][4].Furthermore, space charge distribution results in the distortion of the electric field causing the insulating system to lose efficacy.As a consequence, the malfunction analysis of oil-paper insulation is a complicated process [5].
In order to study the effect of trap characteristics on electric field strength and space charge distribution, considerable research effort has focused on this subject.Ref. [6]  the trap depth of the space charge in a dielectric polymer with molecular dynamics theory and density functional theory.The aging oil demonstrated increased space charge quantity, trap density, and electric field distortion [7].Ref. [8] introduced a theoretical foundation for optimizing the electric field distribution of the insulation system or inhibiting the space charge accumulation.A trap theory model showed the decrease in the space charge quantity in a complex medium [9].The oxide layer broke down when the trap density neared the critical density [10].The aging process of oil was proven to increase the trap levels and trap density [11].Ref. [12] improved the dielectrical property and space charge characteristics by adding montmorillonite (MMT) into epoxy resin and polyethylene.
Montmorillonite (MMT) is an inorganic layered silicate with a natural nano-structure with a large specific surface area and high surface activity [13].In this paper, a novel modified pressboard filled with nano-MMT is developed using the nanocomposite technique to improve the dielectric properties of pressboard.Trap characteristics analyses are performed according to trap theory to examine the reasons for and explore the enhancement mechanism of the breakdown strength and space charge property.

Sample Preparation
The nano doping method in laboratory refers to industrial procreative processes.Raw materials included coniferous kraft pulp, distilled water, and MMT nanoparticles.The process of creating the nano-modified pressboards was divided into six steps: beating, mingle, molding, compressing, desiccation, and immersion oil.The main equipment used for this process are a beater, electronic scales, a pattern forming machine, a curing press, and a vacuum drying oven.The process is described detailed in Figure 1 [14].In order to study the effect of trap characteristics on electric field strength and space charge distribution, considerable research effort has focused on this subject.Ref. [6] built a model to calculate the trap depth of the space charge in a dielectric polymer with molecular dynamics theory and density functional theory.The aging oil demonstrated increased space charge quantity, trap density, and electric field distortion [7].Ref. [8] introduced a theoretical foundation for optimizing the electric field distribution of the insulation system or inhibiting the space charge accumulation.A trap theory model showed the decrease in the space charge quantity in a complex medium [9].The oxide layer broke down when the trap density neared the critical density [10].The aging process of oil was proven to increase the trap levels and trap density [11].Ref. [12] improved the dielectrical property and space charge characteristics by adding montmorillonite (MMT) into epoxy resin and polyethylene.
Montmorillonite (MMT) is an inorganic layered silicate with a natural nano-structure with a large specific surface area and high surface activity [13].In this paper, a novel modified pressboard filled with nano-MMT is developed using the nanocomposite technique to improve the dielectric properties of pressboard.Trap characteristics analyses are performed according to trap theory to examine the reasons for and explore the enhancement mechanism of the breakdown strength and space charge property.

Sample Preparation
The nano doping method in laboratory refers to industrial procreative processes.Raw materials included coniferous kraft pulp, distilled water, and MMT nanoparticles.The process of creating the nano-modified pressboards was divided into six steps: beating, mingle, molding, compressing, desiccation, and immersion oil.The main equipment used for this process are a beater, electronic scales, a pattern forming machine, a curing press, and a vacuum drying oven.The process is described detailed in Figure 1 [14].The finished modified pressboards had a diameter of 200 mm, thickness of 0.40-0.50mm, and a moisture content lower than 0.3%.The microstructure and nanofiller-MMT dispersity of the pressboard nanocomposites was characterized using scanning electron microscopy (SEM) (HITACHI, Tokyo, Japan), and the representative images for neat pressboard and nano-MMT/pressboard composites are shown in Figure 2. The four compared images illustrate MMT nanofillers in 1, 2.5, and 5 wt % filling rates all uniformly dispersed in and composite with the interleaving fiber background without appreciable structural change in the pressboard matrix.The finished modified pressboards had a diameter of 200 mm, thickness of 0.40-0.50mm, and a moisture content lower than 0.3%.The microstructure and nanofiller-MMT dispersity of the pressboard nanocomposites was characterized using scanning electron microscopy (SEM) (HITACHI, Tokyo, Japan), and the representative images for neat pressboard and nano-MMT/pressboard composites are shown in Figure 2. The four compared images illustrate MMT nanofillers in 1, 2.5, and 5 wt % filling rates all uniformly dispersed in and composite with the interleaving fiber background without appreciable structural change in the pressboard matrix.

Measurement System
Conforming to standard ASTM-D149, DC breakdown field strength was measured with cylindrical electrodes.The DC voltage was increased by jogging voltage with a step-up of 2 kV.The quantitative value was calculated from the average of the data to avoid the effect of dispersibility and random error from the equipment and environment.
The measurement system of the thermally stimulated current (TSC) included a Keithley 6517A (America), heating and cooling system vacuum apparatus and a high voltage DC power source.The system and measurement condition are shown in Figures 3 and 4

Measurement System
Conforming to standard ASTM-D149, DC breakdown field strength was measured with cylindrical electrodes.The DC voltage was increased by jogging voltage with a step-up of 2 kV.The quantitative value was calculated from the average of the data to avoid the effect of dispersibility and random error from the equipment and environment.
The measurement system of the thermally stimulated current (TSC) included a Keithley 6517A (America), heating and cooling system vacuum apparatus and a high voltage DC power source.The system and measurement condition are shown in Figures 3 and 4, respectively.

Measurement System
Conforming to standard ASTM-D149, DC breakdown field strength was measured with cylindrical electrodes.The DC voltage was increased by jogging voltage with a step-up of 2 kV.The quantitative value was calculated from the average of the data to avoid the effect of dispersibility and random error from the equipment and environment.
The measurement system of the thermally stimulated current (TSC) included a Keithley 6517A (America), heating and cooling system vacuum apparatus and a high voltage DC power source.The system and measurement condition are shown in Figures 3 and 4  The pulsed electro-acoustic (PEA) measurement system included a high voltage DC source, a signal acquisition system, and a signal recovery system.Two processes occurred during measurement: the reference measurement process and constant voltage measurement process.The electric field stress was 3 kV/mm in the first process, lasting 15 s.The electric field stress was 10 kV/mm and it lasted 3600 s.The measurement system and measurement conditions are shown in Figures 5 and 6, respectively [15].The pulsed electro-acoustic (PEA) measurement system included a high voltage DC source, a signal acquisition system, and a signal recovery system.Two processes occurred during measurement: the reference measurement process and constant voltage measurement process.The electric field stress was 3 kV/mm in the first process, lasting 15 s.The electric field stress was 10 kV/mm and it lasted 3600 s.The measurement system and measurement conditions are shown in Figures 5 and 6, respectively [15].The pulsed electro-acoustic (PEA) measurement system included a high voltage DC source, a signal acquisition system, and a signal recovery system.Two processes occurred during measurement: the reference measurement process and constant voltage measurement process.The electric field stress was 3 kV/mm in the first process, lasting 15 s.The electric field stress was 10 kV/mm and it lasted 3600 s.The measurement system and measurement conditions are shown in Figures 5 and 6, respectively [15].The pulsed electro-acoustic (PEA) measurement system included a high voltage DC source, a signal acquisition system, and a signal recovery system.Two processes occurred during measurement: the reference measurement process and constant voltage measurement process.The electric field stress was 3 kV/mm in the first process, lasting 15 s.The electric field stress was 10 kV/mm and it lasted 3600 s.The measurement system and measurement conditions are shown in Figures 5 and 6, respectively [15].Energies 2018, 11, 1732 5 of 9

Effect of Nano Doping on Thermally Stimulated Current
The TSC spectra of the neat and nanocomposite pressboards were tested and the results are shown in Figure 7 for samples with different MMT filling rates.The TSC spectra for different filling rates depict evident discrepancies in peak current values in the increasing order of I 1.0% < I 5% < I 0% < I 2.5% < I 7.5% and peak temperature positions of T 0% < T 2.5% < T 7.5% < T 5% < T 1.0% .The peak value was 102 pA at 7.5 wt %, 77.6% higher than pure pressboard, which has a value of 57 pA, and 16.52% lower at 1.0 wt % than in the pure sample.

Effect of Nano Doping on Thermally Stimulated Current
The TSC spectra of the neat and nanocomposite pressboards were tested and the results are shown in Figure 7 for samples with different MMT filling rates.The TSC spectra for different filling rates depict evident discrepancies in peak current values in the increasing order of I1.0% < I5% < I0% < I2.5% < I7.5% and peak temperature positions of T0% < T2.5% < T7.5% < T5% < T1.0%.The peak value was 102 pA at 7.5 wt %, 77.6% higher than pure pressboard, which has a value of 57 pA, and 16.52% lower at 1.0 wt % than in the pure sample.

Effect of Trap Characteristics on Electric Breakdown Strength
The DC breakdown electric strength of MMT/pressboard nano-composites with different filling rates are plotted in Figure 8.The electric breakdown fields of the nanocomposite pressboards were remarkably enhanced for mild MMT nanofiller concentrations (representative 1.0 wt % filling rate); nevertheless, when the nano-MMT filling rate increased to higher than 2.5 wt %, the breakdown strength decreased, resulting in an inferior product compared to neat pressboard.

Space Charge Characteristics
The space charge characteristics were measured by PEA and the distribution of the samples with different contents are shown in Figure 9 under an electric field stress of 10 kV/mm.As shown in

Effect of Trap Characteristics on Electric Breakdown Strength
The DC breakdown electric strength of MMT/pressboard nano-composites with different filling rates are plotted in Figure 8.The electric breakdown fields of the nanocomposite pressboards were remarkably enhanced for mild MMT nanofiller concentrations (representative 1.0 wt % filling rate); nevertheless, when the nano-MMT filling rate increased to higher than 2.5 wt %, the breakdown strength decreased, resulting in an inferior product compared to neat pressboard.

Effect of Nano Doping on Thermally Stimulated Current
The TSC spectra of the neat and nanocomposite pressboards were tested and the results are shown in Figure 7 for samples with different MMT filling rates.The TSC spectra for different filling rates depict evident discrepancies in peak current values in the increasing order of I1.0% < I5% < I0% < I2.5% < I7.5% and peak temperature positions of T0% < T2.5% < T7.5% < T5% < T1.0%.The peak value was 102 pA at 7.5 wt %, 77.6% higher than pure pressboard, which has a value of 57 pA, and 16.52% lower at 1.0 wt % than in the pure sample.

Effect of Trap Characteristics on Electric Breakdown Strength
The DC breakdown electric strength of MMT/pressboard nano-composites with different filling rates are plotted in Figure 8.The electric breakdown fields of the nanocomposite pressboards were remarkably enhanced for mild MMT nanofiller concentrations (representative 1.0 wt % filling rate); nevertheless, when the nano-MMT filling rate increased to higher than 2.5 wt %, the breakdown strength decreased, resulting in an inferior product compared to neat pressboard.

Space Charge Characteristics
The space charge characteristics were measured by PEA and the distribution of the samples with different contents are shown in Figure 9 under an electric field stress of 10 kV/mm.As shown in

Space Charge Characteristics
The space charge characteristics were measured by PEA and the distribution of the samples with different contents are shown in Figure 9 under an electric field stress of 10 kV/mm.As shown in Figure 9, the black straight line on the left is the negative electrode and the right light is the positive electrode.Three different nano-modified pressboard contents are displayed including the pure sample, 1 wt % sample, and the 7.5 wt % sample.The 2.5 wt % and 5 wt % samples displayed the similar electric strength and TSC curves as the others.Therefore, the three contents listed above were chosen as objects of our study.
Energies 2018, 11, x FOR PEER REVIEW 6 of 9 Figure 9, the black straight line on the left is the negative electrode and the right light is the positive electrode.Three different nano-modified pressboard contents are displayed including the pure sample, 1 wt % sample, and the 7.5 wt % sample.The 2.5 wt % and 5 wt % samples displayed the similar electric strength and TSC curves as the others.Therefore, the three contents listed above were chosen as objects of our study.

Calculation of Trap Parameters
The trap energy levels and density were calculated using Equations ( 1)-( 3) and analyzing the TSC curve in Figure 7.

ln( )
where Et is tap energy level, k is the Boltzmann constant and , T is temperature, t is time of the rising temperature, v is the vibration frequency calculated with Equation (2), h is Plank constant and 34 6.6 10 J s h     , Nt is the trap density, I is the thermally stimulated current, q is the charge quantity and q = 1 eV, l is the insulated thickness, and f0 is the probability that the trap is occupied by electrons and f0 = 1/2.Figure 10 is a coordinate axis on which the abscissa is the trap energy level and the ordinate is trap density [16][17][18][19]

Calculation of Trap Parameters
The trap energy levels and density were calculated using Equations ( 1)-( 3) and analyzing the TSC curve in Figure 7.
where E t is tap energy level, k is the Boltzmann constant and k = 1.38 × 10 −23 J/K, T is temperature, t is time of the rising temperature, v is the vibration frequency calculated with Equation (2), h is Plank constant and h = 6.6 × 10 −34 J • s, N t is the trap density, I is the thermally stimulated current, q is the charge quantity and q = 1 eV, l is the insulated thickness, and f 0 is the probability that the trap is occupied by electrons and f 0 = 1/2.Figure 10 is a coordinate axis on which the abscissa is the trap energy level and the ordinate is trap density [16][17][18][19].
Energies 2018, 11, x FOR PEER REVIEW 6 of 9 Figure 9, the black straight line on the left is the negative electrode and the right light is the positive electrode.Three different nano-modified pressboard contents are displayed including the pure sample, 1 wt % sample, and the 7.5 wt % sample.The 2.5 wt % and 5 wt % samples displayed the similar electric strength and TSC curves as the others.Therefore, the three contents listed above were chosen as objects of our study.

Calculation of Trap Parameters
The trap energy levels and density were calculated using Equations ( 1)-( 3) and analyzing the TSC curve in Figure 7.

ln( )
where Et is tap energy level, k is the Boltzmann constant and , T is temperature, t is time of the rising temperature, v is the vibration frequency calculated with Equation (2), h is Plank constant and 34 6.6 10 J s h     , Nt is the trap density, I is the thermally stimulated current, q is the charge quantity and q = 1 eV, l is the insulated thickness, and f0 is the probability that the trap is occupied by electrons and f0 = 1/2.Figure 10  According to the calculation result, the order of magnitude of the trap density is 10 21 and the relationship of the maximum value of the trap density with different nanoparticle components is N tm5.0%< N tm1.0%< N tm0% < N tm2.5% < N tm7.5% .The maximum value was 5.81 × 10 21 (1/eV•m 3 ) and the minimum value was 3.06 × 10 21 (1/eV•m 3 ) at 7.5 wt % and 5 wt %, respectively.The trap density was 3.59 × 10 21 (1/eV•m 3 ) in pure pressboard.The relationship of trap energy level with peak value is E t0% < E t7.5% < E t2.5% < E t5.0% < E t1.0% .

Effect of Trap Parameters on Electric Breakdown Strength
The averaged results of the multiple breakdown tests completed for individual different fill rates were calculated and are listed in Table 1.The DC electric breakdown strength of 1.0 wt % MMT/pressboard nano-composite was 218 kV/mm, 16.6% higher than the 187 kV/mm of neat pressboard.However, the breakdown strength distinctly decreased to 158 and 155 kV/mm for 2.5 wt % and 5.0 wt % filling rates, respectively, as the filling rate increased higher than 1.0 wt %, about 17.1% lower than neat pressboard.The charge carriers are in a partially bound state energy level in the energy band gap of the matrix materials.This energy level is capable of capturing charge carriers.The movement of untrammeled charge is directed under the action of the electric field force.These carriers scatter with trap across the trap energy level and they change to a bound state with a certain probability causing the trapping of carriers.Many of these carriers jump into the conduction band creating detrapping carriers.Trapping and detrapping processes are in dynamic equilibrium when the charges release energy during the trapping process and absorb energy in the detrapping process.The process increases trap density and a shallow trap in the 7.5 wt % sample, leading to the higher capacity of catching charges than other samples.The energy of trapping carriers released as ray and heat are transmitted to other trapping carriers.Under the effect of the electric field, many the charge carriers acquire sufficient energy and bonding valence electrons to move to the conduction band of the molecular chain in cellulose, which destroys the chain structure causing electric breakdown.The trap density of pure pressboard was less than that of the 7.5 wt % sample, as the pure pressboard caught fewer charges and released lower energy, resulting in a higher electric breakdown strength.In the 1 wt % sample, the trap had lower density and deeper energy level so it had a higher electric breakdown strength.

Effect of Trap Parameters on Space Charge
When the voltage is first applied, a cathode injects a huge amount of electricity and charge packets with different contents are formed.The charge packet grows on the anode occupying the vast majority of the pressboard over time.The quantity of electric charge is the lowest and the charge packet is smallest in pure pressboards.Compared with the 7.5 wt % sample, the charge packet is longer than with the 1 wt % sample.A large amount of electron holes accumulated at the anodes of the pure and 1 wt % samples, whereas this phenomenon was not obvious in the other samples.
Figure 9 shows the different depths of injecting charges and quantity of electric charge with different nano-modified pressboards.Two sources of space charge are found in dielectric mediums: ionization of impurities and injection of electrode.The former forms a heteropolarity space charge whereas the latter creates a same polarity space charge [20].The distribution curve graphs reveal that the accumulation of heteropolarity charges near the electrode was not obvious, because the concentration of injection was higher than that from ionization.Few charges were injected into specimens and the curves remained constant.

Figure 1 .
Figure 1.The flow diagram follow for creating nano-modified insulating pressboard.

Figure 4 .
Figure 4.The temperature and electric field stress of pressboards varying with time on thermally stimulated current (TSC).

8Figure 4 .
Figure 4.The temperature and electric field stress of pressboards varying with time on thermally stimulated current (TSC).

Figure 6 .
Figure 6.The temperature and electric field stress of pressboards varying with time on the pulsed electro-acoustic (PEA) system.

Figure 8 .
Figure 8.The electric breakdown strength of nano-composite pressboards varying with filling rate.

Figure 8 .
Figure 8.The electric breakdown strength of nano-composite pressboards varying with filling rate.

Figure 8 .
Figure 8.The electric breakdown strength of nano-composite pressboards varying with filling rate.

Figure 10 .
Figure 10.Trap density changing with trap energy level of nano-composite pressboards.
built a model to calculate .
Figure 10.Trap density changing with trap energy level of nano-composite pressboards.Figure 10.Trap density changing with trap energy level of nano-composite pressboards.