- Email: [email protected]
Injection of megavolt electrons into solid dielectrics
Mat. Res. Bull. Vol. 6, pp. 1075-1084, 1971. Pergamon Press, Inc. Printed in the United States.
INJECTION OF MEGAVCLT ELECTROMS INTO SOLID DIELECTRICS John G. Trump and Kenneth A. Wright High Voltage Research laboratory, Mass. Inst. of Tech., Cambridge, Mass.
(Received August 2, 1971) ABSTRACT The phenomena accom~m~ing the injection of megavolt electrons into solid dielectrics and the accumulation and release of trapped electric charge are reviewed. Low intensity injection into Coming 7070 glass at room temperature or meth~imethacrylate at dry ice temperature have produced stable internal fields of several megavolts per cm with average trapped energy densities exceeding (me Joule per cc. Storage times extending from a few hours to years have been terminated either by induced disruptive discharges which leave a permanent pattern in the volume or by thermally controlled release of the stored energy. Pulsed high intensity injection of electrons produces similar though more translent effects on a wider range of dielectric materials. Techniques are described for d ~ t r e t i n g the electrostatic deceleration of injected electrons and for measuring the norms1 and the induced release of the stored energy.
The spectacular electric discharge patterns of the past two centuries (i) etched on dielectric surfaces or in the emulsion of photographic plates exposed to transient overvoltages find their modern counterpart in the intricate 3 dlmensiorml patterns produced within the volume of solid dielectrics on the disrt~tive release of electrons injected by a megavolt particle accelerator.
Earlier investigatorst
working with surfaces in air,
recognized relationships between the magnitude and polarity of the electric stress and the size and delicacy of the resultant pattern.
In the m ~ d l e
193Os yon Hippel (2) extended these observations to films immersed in selected gases at elevated pressures and advanced the scientific explarm-
1075
1076
ELECTRON INJECTION
Vol. 6, No. 10
tion for ~he intricate geometry and detail of these revealing electrical discharge records. The development of Van de Qraaff electrostatic elect~on accelerators at the High Voltage Research Laboratory at M.I.T. at about this time provided ~he opportunity to inject megavolt elec1~ons into the volume of s variety of solid dielectrics. (3)
Certain of these, depending on the
dielectric properties of the msterials and on the ~emperature of the sample, permitted the trapping of electro~,B and the b u i l d - u p of internal eleo%rioal stress which in the limit exceeded the volmne breakdown strength In transparent dielectrics the distributed system of pathways which drained o f f t h e electron-implanted volume i n s few microseconds can be seen as a
permnent visible pattern of co~slderable esthetic and scientific interest. Each shows the character and exquisite individuality suggestive of other great drainage syste~w of nature - the river systems of this earth, the arterial systems of botanical and maw~lian life, or ~he dlsoherges of s~aospheric electricity.
FIG. i shows such a pat~rn instantly vaporized
within s block of luaite on the discharge of the internally trapped electron cloud produced by injection of 3 Hey electrons through its ~ d
sur-
fete. Other studies at ~I.T. with megavolt electrons have dealt (/+,5,6) with their powerful chemical, biological, and physical effects. These lightest and most ubiquitous particles of nature, energized far beyond the few electron-volts characteristic of even the strongest molecular bands, can penetrate signifieantl~ into the volume of a ~ substance to perform
s potentially destructive but often useful function. Polymerization of monomers, cross-linking and stress-relieving of plastics, sterilization of bacterially-contaminated products, inactivation of viruses, the experimental mutation of seeds and organisms, the differential destruction of malignant tissue in humns, the lengthened shelf-life of foods, and the
Vol. 6, No. 10
ELECTRON INJECTION
1077
sterilizatio~ of bone and heart valves for subsequent surgical implant can all be accomplished under appropriate conditions of dosage and control. ~etals, when their temperature rise is restricted,
can withstand a virtually
unlimited irradiation by high energy electrons and thin metal films are usually the "w'L,'~dow' which permit the accelerated beam to pass from vacuum into air.
A portion of the electron energy, particularly when incident on
water-cooled high-atomic-numbered metals,, is converted into far more
FIG. 1 Vaporized discharge pattern produced in lucite after it had been injected with electrons from a 3 Mev Van de Grsaff accelerator
1078
E L E C T R O N INJECTION
Vol. 6, No. 10
penetrating X-rays which produce similar chemical and biological effects. Electron In~ection and Range Injected into dielectrics,
high energy electrons proceed in
their usual way to lose energy largely by inelastic encounters with bound electrons of the molecular structure.
Excitation, ionization, and
scattering to wide angles accompan~ the progress of the primary particle until it is brought to rest.
The scattered trail of each injected electron
is populated with excited and ionized atoms, with occasional delta rays, interstitial a t ~ s displaced by a chance 8tomlc encounter, and the emission of photons whose energy may extend up to the then remaining energy of the injected electron itself.
Since injection of one microcoulomb per sq. ore.
involves nearly 1013 electrons, these basic processes are statistically distributed and randomized.
The maximum penetration of megavolt electrons
into ar~ absorber depends chiefly on absorber density and can be estimated at the rate of one gram per sq. ran. for each 2 Mev of electron energy. Scattering grossly modifies ar~ expected Bragg distribution and displaces the ionization peak to the first third of the electron range.
The measured
distribution of trapped electrons in lucite irradiated by a 3 Mev electron beam is shown in FIG. 2. Though ~rriads of shallow potential wells can he presumed within the 3-dimensionsl atcmic lattice of.most solid dielectrics, few exhibit ability to capture and hold electror~.
Notable among these ere the
borosilicate Coming 7070 glass at room temperature and meth~l~mthacrylate at dry ice temperature.
These have been shown to retain most of the in-
jected charge and can remain electrified for long periods with internal potentials and fields close to the volume b r e a ~ d ~
values.
Other high
resistivity solid dielectrics are capable only of short-time retention, while quartz and Coming 7740 glass do not exhibit significant energy storage possibly because of radiation-induced conductivity during the
Vol. 6, No. 10
E L E C T R O N INJECTION
1079
/,°°-, ¸ g
so
,
so
Intervenin
8 o
(o
0
0.5
1.0
Absorber
1.5
2.0
no
T h i c k n e s s in G m / c m 2 FIG. 2
A. Chsrge distribution for 3 )Aev electrons in lucite. B. Ionization in depth distribution for 3 Mev electrons in al,-.~num.
/ ÷
÷
Conductive coating and ÷ positive surface charge
÷
I--
~
Trapped electron cloud rejected into solid dielectric
-'---- Metal discharger Insulator extension to increase surface flashover strength External load resistance R FIG. 3 Method of d i r e c t i n g d i s r ~ t i v e through sn e x t e r n e l l o e d R.
discharge c u r r e n t s from trapped e l e c t r o n s /
+
+
+
h /
+
i
÷
Me~hod of ~ e r m l l y
÷
I-L
Cooled electrode at T,
+
+
÷
Trapped electron cloud
~
Heated electrode
at Tz
L0ad resistance R FIG. 4
releasing trapped electrum through an external loed R.
1080
ELECTRON INJECTION
Vol. 6, No. 10
injection period. As can be s e e n from FIG. 2 t h e i n i t i a l
distribution
in depth of
trapped e l e c t r o n s does n o t coD/otto w i t h t h e ionlzetion d e n g i t y d i s ~ r i b u t i c n . (5,7) The distribution of ionization determines the chemical end biologlcal effects in the irradiated material, while the dis%rlbutlon of trapped charge determines the internal potential and gradients.
In superior trapping
dielectrics this internal potsntial can exceed the voltage of the electron accelerator.
(9) In air, as eleotro~ injection proceeds, the electric field asso-
ciated with the trapped charge attracts to the external surfaces a corresponding positive stu~aoe charge.
If this surface neutralization has not been com-
pleted by the time the investigator has deactivated the 8ceelerator and reentered the radiation room, a gradually disappearing evidence of external electric field can be no~ed.
This surface charge can be evidenced by dustIDg
the sample and observing the surface discharge patterns produced an triggerLug by a sharp-pointed defect an internal volume discharge.
Ccnduetivel~
costing the dielectric surface and connecting t o a charge-measuring instrumant p r o v i d e s 8 d i r e c t measurement of injected charge b u t does n o t r e c o g n i z e the conductive relaxation which can ooour during end following the injection period. Dose Rate and Lifetime Studies Recent studies by investigators at the Ion Ph~slcs Curperatlon explored the lifetimes of electron-irradiated commercial plastics such
as
lucite, Polyethylene, teflon, and ~ I o n (8). The flat samples, supported in vacuum on 8 temperatu~e-con~olled metal plate, were irradiated at normal incidence from above by 8 2 Mev Van de ~ a f f
accelerator.
The decay of
trapped charge was mecs~red by the ~-field deflection of a i0 Key electron beam passing parallel to add scae distence above the upper sample surface. This work, performed on samples at 0°C or above, gave evidence of a complex ~ o u p of exponential decay constents f o l l o w ~
injection end showed the
Vol. 6, No. 10
ELECTRON
INJECTION
1081
expected d~m~nution of time constant with increase in sample temperature. At low injected doses such as 0.i microcoulomb per cm2 100% of the injected electrons were trapped in Lucite but less ~han 30% in polyet~lene and teflon. The injection of megavolt electrons into solid dielectrics is necessarily modified by the electrostatic repulsion of the internally trapped charge.
This has been demonstrated st Ion Physics by sequential
edgewise photos of the l,-,~nosity produced in the Lucite sheet during inJectien.
These showed the l,-,~nous regions to squeeze progressively
nearer the input surface as the injection advanced.
Calculations of the
electrostatic deceleration of the incoming electrons supported the experimental measurements. These investigators also utilized pulsed Van de Graaff electron sources in studying the dielectric injection phenomena.
Single
pulses of shout 50 nanoseconds duration were found to produce qualitatively similar effects in dielectrics.
Several dielectric materiels, apparently
non-trapping under low intensity injection, were forced by the more than 107 fold increase in injection rate to exhibit a transient storage capability which culminated in the familiar volume discharge pattern. ~Aeasurement and Release of In~ected Energy Calorimetric measurements were made at M.I.T. (I0) of the stored energy of trapped electrons in Co~ning 7070 glass.
These used a simple
calorimeter with 4" of styrofoam insulation around the dielectric disk. energy calibration was made both by measuring the temperature rise of the disk under a known electric heater energy input, end by calculation based on the specific heat and volume of each sample materiel.
The temperature
was measured with a Cu-Constantin thermocouple end a Leeds and Northrup type K potentiometer.
The long time constant of the temperature decay
showed the adequacy of the calorimeter insulation.
The
1082
ELECTRON INJECTION
In the initial measurements,
Vol. 6, No. 10
discharge of the stored energy was
precipitated by impacting the sample with a sharp steel point. punctures were found necessary to discharge the vol~ne fully.
Multiple The tempera-
ture rise of borosilicate glass reached equilibrium in about 5 minutes and could be corrected for the smell calorimetric losses.
Since only a fraction
of the sample was highly charged, these rises averaged below l°C; measured with absolute errors estimated at less then +_ i0 per cent they indicated energy storage densities close to i joule per co. In later studies calorimetric measurement of the stored energy of the electrified dielectric was accomplished under steady discharge induced by slowly raising the sample temperature.
The onset of discharge was
noted by an acceleration in the temperature rise curve.
The added tempera-
ture contribution could be readily translated into its total stored energy. These calorimetric studies suggested the possibility of accomplishing the slow and controlled discharge of the trapped energy by differentially heating the path along which conduction is desired while maintaining the ten,stature in the region of maximum field.
FIG. 3 illustrates
this procedure which would make use of a positive electrode maintained at trapping temperature TI end a negative electrode st an elevated temperature T 2.
As the electrons are released from their traps by this thermal eleva-
tion they are also enabled to flow through the heated dielectric layer and external load R to the positive electrode.
This differential heating method
may permit most of the trapped electrons to flow through ~he iced in intervals controllable from minutes to days by adjustment of T2. FIG. 4 shows a general method of compelling much of the internally stored charge to pass as a high voltage disruptive discharge through a useful iced.
It consists primarily of providing adequate surface flashover
end volume breakdown strength between the selected discharge point and the positive surface charge.
Tests of this method have demonstrated that nearly
Vol. 6, No. 10
ELECTRON INJECTION
1083
all disruptive discharge could be ehannelled through ar~ lo6d resistance R end that reproducible high peek vol~ges of short duration could be thus produced in en external circuit.
1. G. C. Lichtenberg, Novi. Comment. Gott. 8, 168 (1777). 2. F. H. M e r r i l l and A. von Hippel, J. Appl. Phys. 1..O0, 873 (1939). 3. J. G. Trump and R. J. Van de Graaff, J. Appl. Phys. 1_99, 599 (1948). 4. C. G. Dunn, W. L. Campbell, H. F r a m and A. Hutchins, J. Appl. Phys. 1__99, 605 (1948). 5. J. G. Trump, K. A. Wright and A. M. Clark, J. Appl. Phys. 2.1.1, 345 (1950). 6. K. A. Wright and J. G. Trump, J. Appl. Phys. 3...33, 687 (1962). 7. B. Gross and tC A. Wright, Phys. Rev. 114, 725 (1959). 8. J. Dow and S. V. Nablo, Lifetimes of Trapped Charge in Electron Irradiated Dielectrics. IEEE International Conv. Rec., Part 7 (1966). 9. B. Gross and S. V. Nablo, J. Appl. Phys. 38, 2272 (1967). 10. W. W. Chang, Trapping and Discharge of Megavolt Electrons in Solid Dielectrics. M.S. Thesis, EE Dept., M. I. T. (1963).
INJECTION OF MEGAVCLT ELECTROMS INTO SOLID DIELECTRICS John G. Trump and Kenneth A. Wright High Voltage Research laboratory, Mass. Inst. of Tech., Cambridge, Mass.
(Received August 2, 1971) ABSTRACT The phenomena accom~m~ing the injection of megavolt electrons into solid dielectrics and the accumulation and release of trapped electric charge are reviewed. Low intensity injection into Coming 7070 glass at room temperature or meth~imethacrylate at dry ice temperature have produced stable internal fields of several megavolts per cm with average trapped energy densities exceeding (me Joule per cc. Storage times extending from a few hours to years have been terminated either by induced disruptive discharges which leave a permanent pattern in the volume or by thermally controlled release of the stored energy. Pulsed high intensity injection of electrons produces similar though more translent effects on a wider range of dielectric materials. Techniques are described for d ~ t r e t i n g the electrostatic deceleration of injected electrons and for measuring the norms1 and the induced release of the stored energy.
The spectacular electric discharge patterns of the past two centuries (i) etched on dielectric surfaces or in the emulsion of photographic plates exposed to transient overvoltages find their modern counterpart in the intricate 3 dlmensiorml patterns produced within the volume of solid dielectrics on the disrt~tive release of electrons injected by a megavolt particle accelerator.
Earlier investigatorst
working with surfaces in air,
recognized relationships between the magnitude and polarity of the electric stress and the size and delicacy of the resultant pattern.
In the m ~ d l e
193Os yon Hippel (2) extended these observations to films immersed in selected gases at elevated pressures and advanced the scientific explarm-
1075
1076
ELECTRON INJECTION
Vol. 6, No. 10
tion for ~he intricate geometry and detail of these revealing electrical discharge records. The development of Van de Qraaff electrostatic elect~on accelerators at the High Voltage Research Laboratory at M.I.T. at about this time provided ~he opportunity to inject megavolt elec1~ons into the volume of s variety of solid dielectrics. (3)
Certain of these, depending on the
dielectric properties of the msterials and on the ~emperature of the sample, permitted the trapping of electro~,B and the b u i l d - u p of internal eleo%rioal stress which in the limit exceeded the volmne breakdown strength In transparent dielectrics the distributed system of pathways which drained o f f t h e electron-implanted volume i n s few microseconds can be seen as a
permnent visible pattern of co~slderable esthetic and scientific interest. Each shows the character and exquisite individuality suggestive of other great drainage syste~w of nature - the river systems of this earth, the arterial systems of botanical and maw~lian life, or ~he dlsoherges of s~aospheric electricity.
FIG. i shows such a pat~rn instantly vaporized
within s block of luaite on the discharge of the internally trapped electron cloud produced by injection of 3 Hey electrons through its ~ d
sur-
fete. Other studies at ~I.T. with megavolt electrons have dealt (/+,5,6) with their powerful chemical, biological, and physical effects. These lightest and most ubiquitous particles of nature, energized far beyond the few electron-volts characteristic of even the strongest molecular bands, can penetrate signifieantl~ into the volume of a ~ substance to perform
s potentially destructive but often useful function. Polymerization of monomers, cross-linking and stress-relieving of plastics, sterilization of bacterially-contaminated products, inactivation of viruses, the experimental mutation of seeds and organisms, the differential destruction of malignant tissue in humns, the lengthened shelf-life of foods, and the
Vol. 6, No. 10
ELECTRON INJECTION
1077
sterilizatio~ of bone and heart valves for subsequent surgical implant can all be accomplished under appropriate conditions of dosage and control. ~etals, when their temperature rise is restricted,
can withstand a virtually
unlimited irradiation by high energy electrons and thin metal films are usually the "w'L,'~dow' which permit the accelerated beam to pass from vacuum into air.
A portion of the electron energy, particularly when incident on
water-cooled high-atomic-numbered metals,, is converted into far more
FIG. 1 Vaporized discharge pattern produced in lucite after it had been injected with electrons from a 3 Mev Van de Grsaff accelerator
1078
E L E C T R O N INJECTION
Vol. 6, No. 10
penetrating X-rays which produce similar chemical and biological effects. Electron In~ection and Range Injected into dielectrics,
high energy electrons proceed in
their usual way to lose energy largely by inelastic encounters with bound electrons of the molecular structure.
Excitation, ionization, and
scattering to wide angles accompan~ the progress of the primary particle until it is brought to rest.
The scattered trail of each injected electron
is populated with excited and ionized atoms, with occasional delta rays, interstitial a t ~ s displaced by a chance 8tomlc encounter, and the emission of photons whose energy may extend up to the then remaining energy of the injected electron itself.
Since injection of one microcoulomb per sq. ore.
involves nearly 1013 electrons, these basic processes are statistically distributed and randomized.
The maximum penetration of megavolt electrons
into ar~ absorber depends chiefly on absorber density and can be estimated at the rate of one gram per sq. ran. for each 2 Mev of electron energy. Scattering grossly modifies ar~ expected Bragg distribution and displaces the ionization peak to the first third of the electron range.
The measured
distribution of trapped electrons in lucite irradiated by a 3 Mev electron beam is shown in FIG. 2. Though ~rriads of shallow potential wells can he presumed within the 3-dimensionsl atcmic lattice of.most solid dielectrics, few exhibit ability to capture and hold electror~.
Notable among these ere the
borosilicate Coming 7070 glass at room temperature and meth~l~mthacrylate at dry ice temperature.
These have been shown to retain most of the in-
jected charge and can remain electrified for long periods with internal potentials and fields close to the volume b r e a ~ d ~
values.
Other high
resistivity solid dielectrics are capable only of short-time retention, while quartz and Coming 7740 glass do not exhibit significant energy storage possibly because of radiation-induced conductivity during the
Vol. 6, No. 10
E L E C T R O N INJECTION
1079
/,°°-, ¸ g
so
,
so
Intervenin
8 o
(o
0
0.5
1.0
Absorber
1.5
2.0
no
T h i c k n e s s in G m / c m 2 FIG. 2
A. Chsrge distribution for 3 )Aev electrons in lucite. B. Ionization in depth distribution for 3 Mev electrons in al,-.~num.
/ ÷
÷
Conductive coating and ÷ positive surface charge
÷
I--
~
Trapped electron cloud rejected into solid dielectric
-'---- Metal discharger Insulator extension to increase surface flashover strength External load resistance R FIG. 3 Method of d i r e c t i n g d i s r ~ t i v e through sn e x t e r n e l l o e d R.
discharge c u r r e n t s from trapped e l e c t r o n s /
+
+
+
h /
+
i
÷
Me~hod of ~ e r m l l y
÷
I-L
Cooled electrode at T,
+
+
÷
Trapped electron cloud
~
Heated electrode
at Tz
L0ad resistance R FIG. 4
releasing trapped electrum through an external loed R.
1080
ELECTRON INJECTION
Vol. 6, No. 10
injection period. As can be s e e n from FIG. 2 t h e i n i t i a l
distribution
in depth of
trapped e l e c t r o n s does n o t coD/otto w i t h t h e ionlzetion d e n g i t y d i s ~ r i b u t i c n . (5,7) The distribution of ionization determines the chemical end biologlcal effects in the irradiated material, while the dis%rlbutlon of trapped charge determines the internal potential and gradients.
In superior trapping
dielectrics this internal potsntial can exceed the voltage of the electron accelerator.
(9) In air, as eleotro~ injection proceeds, the electric field asso-
ciated with the trapped charge attracts to the external surfaces a corresponding positive stu~aoe charge.
If this surface neutralization has not been com-
pleted by the time the investigator has deactivated the 8ceelerator and reentered the radiation room, a gradually disappearing evidence of external electric field can be no~ed.
This surface charge can be evidenced by dustIDg
the sample and observing the surface discharge patterns produced an triggerLug by a sharp-pointed defect an internal volume discharge.
Ccnduetivel~
costing the dielectric surface and connecting t o a charge-measuring instrumant p r o v i d e s 8 d i r e c t measurement of injected charge b u t does n o t r e c o g n i z e the conductive relaxation which can ooour during end following the injection period. Dose Rate and Lifetime Studies Recent studies by investigators at the Ion Ph~slcs Curperatlon explored the lifetimes of electron-irradiated commercial plastics such
as
lucite, Polyethylene, teflon, and ~ I o n (8). The flat samples, supported in vacuum on 8 temperatu~e-con~olled metal plate, were irradiated at normal incidence from above by 8 2 Mev Van de ~ a f f
accelerator.
The decay of
trapped charge was mecs~red by the ~-field deflection of a i0 Key electron beam passing parallel to add scae distence above the upper sample surface. This work, performed on samples at 0°C or above, gave evidence of a complex ~ o u p of exponential decay constents f o l l o w ~
injection end showed the
Vol. 6, No. 10
ELECTRON
INJECTION
1081
expected d~m~nution of time constant with increase in sample temperature. At low injected doses such as 0.i microcoulomb per cm2 100% of the injected electrons were trapped in Lucite but less ~han 30% in polyet~lene and teflon. The injection of megavolt electrons into solid dielectrics is necessarily modified by the electrostatic repulsion of the internally trapped charge.
This has been demonstrated st Ion Physics by sequential
edgewise photos of the l,-,~nosity produced in the Lucite sheet during inJectien.
These showed the l,-,~nous regions to squeeze progressively
nearer the input surface as the injection advanced.
Calculations of the
electrostatic deceleration of the incoming electrons supported the experimental measurements. These investigators also utilized pulsed Van de Graaff electron sources in studying the dielectric injection phenomena.
Single
pulses of shout 50 nanoseconds duration were found to produce qualitatively similar effects in dielectrics.
Several dielectric materiels, apparently
non-trapping under low intensity injection, were forced by the more than 107 fold increase in injection rate to exhibit a transient storage capability which culminated in the familiar volume discharge pattern. ~Aeasurement and Release of In~ected Energy Calorimetric measurements were made at M.I.T. (I0) of the stored energy of trapped electrons in Co~ning 7070 glass.
These used a simple
calorimeter with 4" of styrofoam insulation around the dielectric disk. energy calibration was made both by measuring the temperature rise of the disk under a known electric heater energy input, end by calculation based on the specific heat and volume of each sample materiel.
The temperature
was measured with a Cu-Constantin thermocouple end a Leeds and Northrup type K potentiometer.
The long time constant of the temperature decay
showed the adequacy of the calorimeter insulation.
The
1082
ELECTRON INJECTION
In the initial measurements,
Vol. 6, No. 10
discharge of the stored energy was
precipitated by impacting the sample with a sharp steel point. punctures were found necessary to discharge the vol~ne fully.
Multiple The tempera-
ture rise of borosilicate glass reached equilibrium in about 5 minutes and could be corrected for the smell calorimetric losses.
Since only a fraction
of the sample was highly charged, these rises averaged below l°C; measured with absolute errors estimated at less then +_ i0 per cent they indicated energy storage densities close to i joule per co. In later studies calorimetric measurement of the stored energy of the electrified dielectric was accomplished under steady discharge induced by slowly raising the sample temperature.
The onset of discharge was
noted by an acceleration in the temperature rise curve.
The added tempera-
ture contribution could be readily translated into its total stored energy. These calorimetric studies suggested the possibility of accomplishing the slow and controlled discharge of the trapped energy by differentially heating the path along which conduction is desired while maintaining the ten,stature in the region of maximum field.
FIG. 3 illustrates
this procedure which would make use of a positive electrode maintained at trapping temperature TI end a negative electrode st an elevated temperature T 2.
As the electrons are released from their traps by this thermal eleva-
tion they are also enabled to flow through the heated dielectric layer and external load R to the positive electrode.
This differential heating method
may permit most of the trapped electrons to flow through ~he iced in intervals controllable from minutes to days by adjustment of T2. FIG. 4 shows a general method of compelling much of the internally stored charge to pass as a high voltage disruptive discharge through a useful iced.
It consists primarily of providing adequate surface flashover
end volume breakdown strength between the selected discharge point and the positive surface charge.
Tests of this method have demonstrated that nearly
Vol. 6, No. 10
ELECTRON INJECTION
1083
all disruptive discharge could be ehannelled through ar~ lo6d resistance R end that reproducible high peek vol~ges of short duration could be thus produced in en external circuit.
1. G. C. Lichtenberg, Novi. Comment. Gott. 8, 168 (1777). 2. F. H. M e r r i l l and A. von Hippel, J. Appl. Phys. 1..O0, 873 (1939). 3. J. G. Trump and R. J. Van de Graaff, J. Appl. Phys. 1_99, 599 (1948). 4. C. G. Dunn, W. L. Campbell, H. F r a m and A. Hutchins, J. Appl. Phys. 1__99, 605 (1948). 5. J. G. Trump, K. A. Wright and A. M. Clark, J. Appl. Phys. 2.1.1, 345 (1950). 6. K. A. Wright and J. G. Trump, J. Appl. Phys. 3...33, 687 (1962). 7. B. Gross and tC A. Wright, Phys. Rev. 114, 725 (1959). 8. J. Dow and S. V. Nablo, Lifetimes of Trapped Charge in Electron Irradiated Dielectrics. IEEE International Conv. Rec., Part 7 (1966). 9. B. Gross and S. V. Nablo, J. Appl. Phys. 38, 2272 (1967). 10. W. W. Chang, Trapping and Discharge of Megavolt Electrons in Solid Dielectrics. M.S. Thesis, EE Dept., M. I. T. (1963).