AD-A253 850 DTIChhhhh~l~flllhl S ELECT E

JUL 17 19921MONOLITHIC INTEGRATION

cr CSEMICONDUCTOR AND SUPERCONDUCTOR COMPONENTS

DARPA/ONR Contract No. N00014-90-C-0226

Honeywell Sensor and System Development Center10701 Lyndale Avenue South

Bloomington, MN 55420

1 April 1992 -30 June 1992

A p v- ( T-., ' f r P t !I A1 ....____d 92-18644

I.l Ii I II i i miii92714 0,06

Program Manager. Paul W. Krum Te IAJ W 4mm Phone #: 612M87-4441

Principal Investigator. Busa Johnson hone #: 612/887-4505

Section Head: Phone - 612.887-4576

Department Norma A. Fos Phone#-.: 612W87-6424

Pape I ~.&qa~J

2.0 PROGRAM SUMMARY

The goal of the program is to develop transistor technology compatible withhigh transition temperature superconductor technology so that transistorpixel switches can be integrated with YBa2Cu 307 superconductingmicrobolometers in the same silicon substrate. A 4x4 matrix-addressablesuperconducting microbolometer array will be delivered at the completionof the program.

&.0 PROGRAM STATUS

Task 1.0: Vendor Selection

Mary Weybright, a graduate student in electrical engineering at StanfordUniversity working under the direction of Prof. James D. Plummer, hasbeen engaged as a consultant to model the performance of bipolartransistors at low temperature, in order to determine doping profilesneeded for the transistor switches at each pixel.

The monolithic bipolar transistors have been fabricated by Honeywell'sMICRO SWITCH Division of Richardson, Texas, and donated to thecontract at no cost to the contract. The transistor fabrication task has beendeleted from the contract statement of work.

Task 2.0: First Fabrication Run (completed)

Task 2.1: Film Development

Superconducting films of YBa 2Cu307 were grownin-situ on 3 - inch silicon wafers coated withamorphous silicon nitride and polycrystallineyttria stabilized zirconia (YSZ). The films weregrown by ion beam sputtering. These films showonset of superconductivity at -88 K and zeroresistance at -65 K The temperature coefficient ofresistance (TCR) was about 0.15 K-1 at themidpoint of the transition (-73 K). The substratesused are ideal for fabrication of microstructuresby silicon micromachining techniques. Thesefilms were grown using a combination of pureozone from Honeywell's ozone distillation system,and ordinary oxygen. The growth temperature foroptimum superconducting properties was - -/

between 700* C and 7350 C. These growth ....temperatures are believed to be low enough to >allow survival of the transistors which will be atembedded in the substrate. &0:-u ,

A"allabil ity oPage 2 T~n5mAp.u and/a;Dist Special

DTIC QUALTY INC'2ECTED ... [ .

Task 2M

The first fabrication run used existing masksfrom DARPA/ONR Contract #N00014-88-C-0394.

Task 2A Vendor Electronics

For the first fabrication run, transistors were notembedded in the substrate.

Task 2A4: Intead Device

No working bolometers were produced in the firstfabrication run. Due to inadequate passivatiol,the YBa2Cu307 film on the bolometer structureswas attacked by the KOH solution that was used toetch the silicon to thermally isolate thebolometers. In addition, the electrical contactpads did not adhere to the substrate. Theseproblems were solved in a subsequent processingrun funded by Honeywell which was completed inDecember, 1991.

Task 2.5: Device Evaation

Room temperature resistance measurementswere performed on the first two wafers of the firstfabrication run.

Task 3.0:. Second Fabrication Run

Task 3.1: Film Develm nt

The superconducting properties of the YBa2Cu307films on silicon wafers coated with Si3N4 and YSZhave been improved compared with those grownunder task 2.1. The oxygen pressure in thechamber has been increased and a small amountof silver is co-sputtered with the Y, Ba, and Cu inorder to increase the critical current and reducethe noise in the films. Some films show onset ofsuperconductivity at -88 K and zero resistance at-72 K, with a temperature coefficient of resistance(TCR) of about 0.30 K-1 at the midpoint of thetransition (-79 K). A film with this TCR on athermally isolated microstructure could provide abolometer with a sensitivity which is high enoughfor high performance imaging applications usinga staring array. Other films have been grownwith a transistion midpoint temperature of 83 K,

Pape 8 3mawp~uu

but the TCR of these films at the transitionmidpoint is only about 0.16 K-1. The substratesused are ideal for fabrication of microstructuresby silicon micromachining techniques. Thegrowth temperature for optimumsuperconducting properties (720°C to 740 0C) is stillbelieved to be low enough to allow survival of thetransistors which will be embedded in thesubstrate.

The 1/f noise in YBa 2Cu30 7 fims grown atHoneywell on silicon substrates with Si3N4/YSZbuffer layers is now low enough that if they wereused in a large thermal imaging array ofmicrobolometer pixels 50 pm x 50 pm in size, the1/f noise voltage would be only about 4 times largerthan the Johnson noise voltage, under typicaloperating condtions.

Noise measurements have been performed at theUniversity of Minnesota, our subcontractor, onDyBa2Cu307 films grown at the University ofMinnesota on SrTi0 3 and LaA103 single crystalsubstrates. Measurements of the peak in theexcess noise spectral density at the foot of theresistive transition in these films show that forlow magnetic fields (less than approximately 0.5Tesla), the excess noise is due to vortex motion.The noise does not appear to be caused by chargetrapping and detrapping at grain boundaries. Formagnetic fields larger than about 0.5 Tesla, thenoise properties are qualitatively different. This isthought to be due to the density of vorticesbecoming comparable to the density of grainboundary junctions in the samples.

Task 3.2: Mak D ig

The masks for the transistors andmicrobolometers have been delivered. In the 4 x 4arrays, each pixel (including transistor) willoccupy an area 125 pm x 200 pm.

Task 3.3: Vendor ectronics

Honeywell's MICRO SWITCH Division hasdelivered the wafers with bipolar transistors thatwill be used in the second fabrication run toproduce 4x4 arrays of microbolometers with atransistor at each pixel. These transistors werefabricated by modifying MICRO SWITCH'sstandard fabrication process according to

Page 4 WL8M.I*39AP"U

suggestions by Mary Weybright, our consultant atStanford University, in order to achieve goodperformance at low temperatures. MaryWeybright's suggestions were guided by previouslow temperature (77 K) measurements ontransistors from a pilot run using MICROSWITCH's standard process.

Task 3.4: Inte ted Device

Experiments have been carried out to determinethe microbolometer process steps that causesignificant damage to the transistors. Subjectingthe transistors to 750*C in a furnace in nitrogenfor 4 hours (in order to simulate the conditions ofYBa2Cu 307 film growth) causes a reduction byabout a factor of 2 in the room temperaturecurrent gain for some transistors. However, fortransistors fabricated slightly differently, therewas minimal damage after this heat treatment.The sputter deposition of the Si3N 4 layer thatprovides the underside of the thermally isolatedstructures causes a reduction in the current gainof the transistors, but the damage is easilyreversed with a rapid thermal anneal in air at410*C for 10 seconds. Annealing in forming gas,the preferred method of repairing transistordamage caused by sputtering, is not possiblebecause the hydrogen in the forming gas convertsthe YBa 2Cu30 7 film into an insulator. Other teststo determine damage to the transistors duringmicrobolometer processing are in progress.

The microbolometer process sequence for thesecond fabrication run has been established,except for possible modifications to avoidprocessing damage to the transistors.

Task 35: Device Ealuan

Measurements of transistor performance at lowtemperature show a current gain of 20 at 77 KThis is believed to be an adequate current gain foroperation of a large focal plane array ofmicrobolometers with transistors at each pixel.

All 24 of the wafers delivered by Honeywell'sMICRO SWITCH Division have beencharacterized at room temperature. The currentgain and leakage current were measured.

Fag. 5

Devices fabricated in the recent processing runfunded by Honeywell have been evaluated. Aresponsivity of 1300 volts/watt with a dc biascurrent of only 1 pA was measured at a substratetemperature of 73 K in a microbolometeroccupying a 125 um x 125 un area. In thermalimaging applications, pulsed bias currents ofabout 100 pA would be used, resulting in aresponsivity of 130,000 volts/watt. A 12-elementlinear array showed responsivity nonuniformityless than 7% over most of the array. Thedetectivity, D*, measured at a frequency of 7 Hz,was 7.5 x108 cm Hz1 2/Watt. The noise wasdominated by 1/f noise in the Au/YBa 2Cu307contacts. It is expected that this contact noise canbe eliminated by straightforward processmodifications, resulting in an improvement inthe D* at 7 Hz by about a factor of 10. In themicrobolometers that were tested, there was nosignificant degradation of the electrical propertiesof the superconductor during the processingrequired to fabricate the microbolometers.

Task 4.0: Third Fabrication Run

A request has been submitted to delete this task from thecontract statement of work, due to lack of funds.

4.0 ACCOMPLIMENTS (for 1 April 1992 to 30 June 1992)

Task 1.0: Vendor Selection

No work was performed on this task during this period.

Task 2.0: First Fabrication Run

No work was performed on this task during this period.

Task 3.- Second Fabrication Run

Task 31: Film Development

The resistive transition of the films grown atHoneywell on amorphous Si3N4 with apolycrystalline YSZ buffer layer has beenimproved by introducing a small amount of watervapor to the deposition chamber during filmgrowth. A film was grown in this way withsuperconducting onset temperature of 91 K, atransition midpoint temperature of 83 K andzero resistance at a temperature of 70 K

Page 6 a~mm&upsrWpN.1u

Task Di

The mask design for the microbolometerfabrication was completed and the masks weredelivered. In the 4 x 4 arrays, each pixel(including transistor) occupies an area 125 pm x200 pgm.

Task 13: Vendor Electronics

The monolithic bipolar transistors fabricated byHoneywell's MICRO SWITCH Division ofRichardson, Texas were delivered. Thesetransistors were donated to the contract at no costto the contract. The transistor fabrication taskhas been deleted from the contract statement ofwork.

Task 34: Integated Device

Experiments were carried out to determine themicrobolometer process steps that causesignificant damage to the transistors. Subjectingthe transistors to 7500C in a furnace in nitrogenfor 4 hours (in order to simulate the conditions ofYBa2Cu307 film growth) causes a reduction byabout a factor of 2 in the room temperaturecurrent gain for some transistors. However, fortransistors fabricated slightly differently, therewas minimal damage after this heat treatment.The sputter deposition of the Si3N4 layer thatprovides the underside of the thermally isolatedstructures causes a reduction in the current gainof the transistors, but the damage is easilyreversed with a rapid thermal anneal in air at4100 C for 10 seconds. Annealing in forming gas,the preferred method of repairing transistordamage caused by sputtering, is not possiblebecause the hydrogen in the forming gas convertsthe YBa2Cu307 film into an insulator. Other testson microbolometer processing damage to thetransistors are in progress.

The microbolometer process sequence for thesecond fabrication run was established, except forpossible modifications to avoid processing damageto the transistors.

Page 7

Task 35: Device Evauation

Measurements of transistor performance at lowtemperature show a current gain of 20 at 77 KThis is believed to be an adequate current gain foroperation of a large focal plane array ofmicrobolometers with transistors at each pixel.

All 24 of the wafers delivered by Honeywell'sMICRO SWITCH Division were characterized atroom temperature. The current gain and leakagecurrent were measured.

Task 4.0 Third Fabrication Run

A request has been submitted to delete this task from thecontract statement of work, due to lack of funds.

L0 PROBLEM AREAS49S&UES

* Performance of transistors at low temperature must be adequate forgood switching performance.

* Transistors must survive the high growth temperature of theYBa2Cu307 films, as well as the plasma and ion beam presentduring YBa2Cu307 film growth.

" Transistors must survive the microbolometer processing steps, suchas sputter deposition of films, ion milling, and PECVD deposition ofSi3N4.

" The electrical resistance of the gold contacts to the YBa2Cu30 7 filmsmust be low enough that small area (-10 um x 10 gim) contacts have aresistance low compared to the bolometer resistance. Small areacontacts are necessary for high fill-factor in two-dimensional arraysof microbolometers.

* The noise generated in the gold contacts to the YBa 2Cu307 must beow compared to the noise in the YBa2Cu307 films.

60 CORRECTIVE ACION

" Utilize calculations of transistor performance to optimize lowtemperature performance.

" Reduce the growth temperature of the YBa2Cu307 films as much aspossible.

" Use PtSi ohmic contacts to the transistors.

P*P 8 ~it8~.~~g

* Substitute sputtered Si3N4 passivation for the PECVD depositedSi3N4, in order to minimize damage to the transistors.

* Reduce electrical contact resistance and contact noise by depositinggold in-situ by ion beam sputtering immediately after deposition ofthe YBa2Cu307 film. The gold film will then be photolithographicallypatterned into the desired contact geometry.

7.0 GOALS FOR THE NEXT PERIOD (1 July 1992 to 30 Sepmbe 1992).

* Complete a process run to fabricate 4 x 4 arrays of microbolometersintegrated with transistors.

* Evaluate the 4 x 4 arrays of microbolometers integrated withtransistors.

" Write the final report.

&0 PUBLICATIONS

8.1 E Mn Pubhadin'd onata

None

8,2 EBB= l~l~ in Conferen ce ed"nMAIM

B.R. Johnson, T. Ohnstein, H. Marsh, S.B. Dunham and P.W.Kruse, "YBa2Cu307 Superconducting Microbolometer LinearArrays," to be published in SPIE Conference Proceedings volume1685, Infrared Detectors and Focal Plane Arrays.

B.R. Johnson, C-J Han, R.E. Higashi, P.W. Kruse and R.A. Wood,"Monolithic Silicon Focal Planes with High-Te Superconductor IRSensors," to be published in the Proceedings of the IRIS SpecialtyGroup Meeting on Infrared Detectors, NASA Ames, August 18-21,1992.

a. Invitd

B.R. Johnson, "Superconducting Microbolometer InfraredDetector Arrays on Silicon Microstructures," presented at theInternational Superconductor Applications Convention, SanDiego, CA, January 14-16, 1991.

P. W. Kruse, "Focal Plane Arrays Based Upon ThermalDetection Mechanisms," JASON Imaging Infrared DetectorWorkshop, LaJolla, CA, June 29, 1992.Paul W. Kruse, "Focal Plane Arrays Based Upon Thermal

Detection Mechanisms," Defense Sciences Research Council

Page 9 ML..&8SAVW4=

Workshop on Multispectral Infrared Systems, LaJolla, CA,July 14, 1992.

B.R. Johnson, T.R. Ohnstein, C-J Han, R.E. Higashi, P.W.Kruse, RA. Wood, HA. Marsh, and S.B. Dunham, "High TeSuperconducting Microbolometer Arrays Fabricated by SiliconMicromachining," to be presented at the AppliedSuperconductivity Conference, Chicago, IL, August 24-28,1992.

b. C nriuated

B.R. Johnson, C-J Han, T.R. Ohnstein, B.E. Cole and P.W.Kruse, "Monolithic Integration of Semiconductor andSuperconductor Components," DARPA Second Annual HighTemperature Superconductors Workshop, Sheraton Tara Hoteland Resort, Danvers, MA, October 3-5, 1990.

B.R. Johnson, T.R. Ohnstein, P.W. Kruse and S.B. Dunham,"YBa2Cu307 Films for Infrared Bolometers on SiliconMicrostructures," DARPA Third Annual High TemperatureSuperconductors Workshop, Hyatt Bellevue Hotel, Bellevue,Washington, September 30 - October 2, 1991.

B.R. Johnson, P.W. Kruse, and S.B. Dunham, "YBa2Cu307Films For Infrared Bolometers on Silicon Microstructures,"Materials Research Society Fall Meeting, Boston, MA,December 2-6, 1991.

P.W. Kruse, "Fundamental Limits of Infrared Detectors andArrays," JPL Innovative Long Wavelength Detector Workshop,Pasadena, CA, April 7-9, 1992.

B.R. Johnson, T.R. Ohnstein and P.W. Kruse, "High TcSuperconducting Microbolometer Linear Arrays," JPLInnovative Long Wavelength Detector Workshop, Pasadena,CA, April 7-9, 1992.

B.R. Johnson, P.W. Kruse, T.R. Ohnstein, C-J Han and R.E.Higashi, "High Tc Superconductor Infrared Bolometers onSilicon Microstructures," American Ceramic Society Meeting,Minneapolis, MN, April 13-17, 1992.

B.R. Johnson, T.R. Ohnstein, H.A. Marsh, S.B. Dunham andP.W. Kruse, "YBa2Cu307 Superconducting MicrobolometerLinear Arrays," SPIE OE/Aerospace Sensing Meeting,Orlando, FL, April 20-24, 1992.

B.R. Johnson, C-J Han, R.E. Higashi, P.W. Kruse and R.A.Wood, "Monolithic Silicon Focal Planes with High-TeSuperconductor IR Sensors," to be presented at the IRIS

Pag 10 Mwbkq.rA.,ej, U

Specialty Group on Infrared Detectors Meeting, NASA Ames,Moffett Field, CA, August 18-21, 1992.

B.R. Johnson, P.W. Kruse, S.B. Dunhamn, H.A. Marsh, C.J.Han and R.E. Higashi, "YBa2Cu3O7 . Microbolometer ArraysFabricated by Silicon Micromachining," Materials ResearchSociety Fall Meeting, Boston, MA, November 30-December 4,1992.

9.0 FINANCIAL

A. Funding Authorized $56Z022

B. Funds Expended or Committed $609,512(Week ending 30 June 1992)

C. Additional Funds Required from ONR $89,271to Complete Contract, Assuming ThirdFabrication Run Is Deleted fromStatement of Work

Page I1 IIiI8~p~a

1" z0

a

a

u

0

IMa, 4

|z.

b UU

",uW2