Porous Silicon For Pumping Solid-State Lasers: Ii - LC L LLLLLLLLL
Porous Silicon For Pumping Solid-State Lasers: Ii - LC L LLLLLLLLL
S. D. Russell
W. B. Dubbelday
R. L. Shimabukuro
2,9993
I 93-14885
Ii•lc>l~lllllllll
ISLJ(LAIMEIt NOTICE
S. D. Russell
W. B. Dubbelday
R. L. Shimabukuro
NAVAL COMMAND, CONTROL AND
OCEAN SURVEILLANCE CENTER
RDT&E DIVISION
San Diego, California 92152-5001
J. D. FONTANA, CAPT, USN R.-T. SHIEARER
Commanding Officer Executive Director
ADMINISTRATIVE INFORMATION
The research presented in this work was funded by the Office of Naval Rescarch.
Solid-State Laser (6.1.5) block program and in part by the NCCOSC. RDT&E Division
Independent Research (IR) program.
ACKNOWLEDGMENTS
The authors thank Dr. E Hanson (NRaD. Code 843) and Dr. M. Sailor (Dept. of
Chemistry, UCSD) for making their labs available for the collection of photoemission
spectra, and S. Pappert (NRaD, Code 555) foi supplying the multiple quantum well Al-
GaAs/GaAs sample used in this investigation. The authors also thank
D. Gookin (NRaD, Code 843) for exchanging information concerning a novel applica-
tion for porous silicon, and Dr. V. Smiley (NRaD, Code 804) and Dr. A. Gordon
(NRaD, Code 0014) for funding this research. Finally, the authors thank Dr.
D. Szaflarski (ONT/ASEE postdoctoral fellow, now at Molecular Biosystems) and
P. Georgief (SDSU Foundation Student) for preparing some of the samples used
in this work.
SM
CONTENTS
BACKGROUND ..................................................
DIODE-PUMPED SOLID-STATE LASERS ................... .
POROUS SILICON .............................................. 3
O BJE CT IV E ................................................... ........ 4
APPROACH .................................................... ... 4
D E LIV E RA BLES ..................................................... 4
PERIOD OF PERFORMANCE ....................................... 4
EXPERIM ENTAL RESULTS .............................................. 4
POROUS SILICON FABRICATION PROCESS ........................... 4
POROUS SILICON ON SAPPHIRE ..................................
QUANTUM EFFICIENCY OF POROUS SILICON ....................... 7
RECENT RESU LTS .................................................. If
FUTURE DEVELOPMENTS ........ ................................ 0
CO N C LU SIO N S ........................................................ I
R E FE R E N C E S ......................................................... I
FIGURES
1. Common energy levels for solid-state lasers . ............................... 2
NT:, FPO
-u:,t'......
I -
iii
BACKGROUND
Solid-state lasers have military, scientific, and commercial applications in. for example.
ranging, imaging, spectroscopy, nonlinear optics, and materials processing. Diode laser pump-
ing offers an all-solid-state laser with high efficiency in a simpler. more reliable, and compacl
system. These systems offer many improvements over the use of diode lasers themselves, in-
cluding narrower frequency linewidth. better beam quality. higher peak powers, and different
wavelengths. Diode lasers are attractive as pumps because they are compact, efficient, and re-
liable. However, since they are fabricated from III-V compound semiconductors. they are ex-
pensive. The emission wavelengths of diode lasers may be adjusted by varying the material
composition during wafer growth and precisely matched to coincide to peak absorption lines
by controlling the operating temperature. Therefore, diode pumps cause less undesirable
heating and produce no damage such as may occur when high-energy photons from flashlamps
are present in the pump light. In addition, because the diode pumps are partially coherent de-
vices, the output beams may be efficiently focused and optically coupled to spatially match the
solid-state laser mode.
Solid-state lasers operate in various manners and configurations, and virtually all of these
lend themselves to the use of diode pumps. The geometrical configurations include a rigid
cylindrical rod, a slab, and a flexible optical fiber. Solid-state lasers operate in CW, long-pulse
(jts time frame) and short-pulse (0-switched) forms. The emission may contain single or multi-
ple spatial or longitudinal modes. The use of diode pumps provides for several new pump-
resonator configurations; for example, transversely pumped rod or slab systems for pulsed
high-power multimode output, analogous to flashlamp-pumped systems, longitudinally CW
end-pumped rod systems for low-power output with high efficiency because of the ability for
optical mode matching in the lasing volume; and longitudinally pumped scalable slab or disc
systems for high CW power in the fundamental (TEM 0 0) mode by increasing both the radial
and longitudinal component of the pumped volume [1].
Diodes fabricated from AIGaAs with different concentrations of Al and Ga emit at wave-
lengths of about 800 nm, ranging from about 700 to 900 nm. These emission wavelengths coin-
cide well with the strong absorption bands of several currently important solid-state lasing
ions, including, for example, neodymium, holmium, erbium, and promethium. Typical pump-
ing requirements at the desired wavelength are tens of milliwatts to hundreds of milliwatts.
Threshold pumping values of 6-12 mW for Nd:NPP (Nd:P 5 0 14) and LNP (LiNdP 4 O; 2) with
reasonable efficiency have been reported [2]. Laser emission has also been reported from
Ho:YAG at 2.1 Rm, Er3+ : YLFat 2.8 Rtm, Nd:YAG at 0.946 Um, Nd : YVO 4 -at 1.06 .am. and
Nd:glass at 1.06 rim; these and others have been thoroughly reviewed in the scientific literature
[I -41. Figure 1 describes some common energy level schemes for solid-state lasing, and the
desired pump wavelengths [2]. This shows that pump sources emitting between 700 and 900 nm
may be employed to pump a large variety of laser hosts. Additional information can be ob-
t-i '7ed from,! KammsKHi [4J. I he host crystal tor these lasing ions plays a iole in Jie ime structure
of the absorption bands due to the Stark splitting by the crystal field, but does not substantially
change the required spectral region required for the efficient pumping of the ions into the
excited states desired for laser emission. For reference, the best optical efficiency in dil)de-
pumped solid-state lasers is 197' (Nd:YAG) [3] compared to 8t7,4 efficiency in a flashlamp-
pumped coaxially cylindrical Nd:glass rod, although more typically it is 51'4 with a flashlamp
with Nd,Cr:GSGG or Er, Tm, Ho:YAG [2]. Therefore, research into high-cfficiency pumping
schemes is ongoing.
112
i: 5
j _,j I
ci • •,.--÷ at
hnes TheI'irst
popo,,ed ioep ' pd
a .. lkn)lo i•"'¢:4. :[•S2•
ner
0 ,rnpur~e n Er8g tr •pllp•;O h ,r.natv
OW
l n2 is
aa
in
5- a1 •"
it + i ,
2
POROUS SILICON
* Private communication from M. J. Sailor. Dept. of Chemistry, University of California. San Diego (UCSI)).
3
OBJECTIVE
The objective of this research is to perform pre'iminarv experiments to examine the feasi-
bility of using porous silicon and related devices for pumping solid-state lasers.
APPROACH
The controlled fabrication of light-emitting silicon structures will be developed using
chemical, electrochemical, or photochemical techniques. Emphasis will be placed on tech-
niques compatible with the requirements for the VLSI device processing of wafer size and
etching uniformity. The optical and microstructural properties will be characterized using in-
house facilities. This will elucidate the emission mechanism, which may allow the required
engineering of the emission wavelength peak. In addition, variations in processing conditions
and in silicon film thickness on sapphire will be examined for more precise control of the
porous silicon thickpess. These methods may provide the means to control the peak wave-
length of the emission and may also allow control of the emitted linewidth. The photolumincs-
cence intensity from porous silicon will be quantitatively measured for comparison with com-
peting semiconductor pump sources.
DELIVERABLES
A preliminary report summarizing the experiments performed and conclusions was sub-
mitted 30 September 1992. The current technical document is the final deliverable and will
complete this preliminary study of the feasibility of using porous silicon as a solid-state laser
pump source.
PERIOD OF PERFORMANCE
The period of performance for this work was 1 June 1992 to 30 September 1992.
EXPERIMENTAL RESULTS
4
investigation be conducted of fabrication techniquleS more compatible wiih conmentioln:l
VLSI processing. In this work, we demonstrated the first photochcmical fa rlcattion of porous
silicon, and also investigated chemical stain etches rpeportcd in the literature [14.151 as requir-
ing no electrical current.
The chemical stain etch used a solution of HF:FINO 3 :dcionized ft1O (1:5: P(1). Bulk silic0n
and SOS wafers were similarly processed. The bulk wafers were (100)-oriented B-doped ,i1i-
con with 3-5 Q-cm resistivity. The SOS samples were cpitaxially deposited silicon, boron-
doped in situ to 4 x 10 15 cm-, on 270-nm-thick SOS (1 X lt0I' cm-3) to a total silicorn
thickness of 10 lim. The etch solutions were prepared by reacting a square centimeter of silicon
with the HF:HNO 3 mixture for 2 minutes, causing an accumulation of tItNO,. the activc
oxidizer in the reaction [171. The chemical etch is a result of hole injection from the tNO(
oxidant into the silicon substrate via the following net reaction ! 161:
Si + HNO 3 + 6 HF --- H 2 SiFe + HNO 2 + H2 0 + H';
The reaction is catalyzed by the presence of NO 2 ions, and so there is usually an induction
period observed for the etching process. Deionized H_,O was subsequently added to the solu-
tion before the sample to be etched was immersed. Typical etch times ranged from I to 15 min-
utes. The samples were rinsed with deionized water, dried with nitrogen, and examined with a
hand-held ultraviolet lamp (Mineralight Model #S52). Generally, samples etched for less
than 1 minute did not luminesce, whereas SOS samples that are etched longer than - 15 min-
utes resulted in the complete dissolution of the silicon off the sapphire substrate.
We also demonstrated for the first time that photoluminescent porous sillcon is produced
via a photoinitiated chemical stain etch. For these experiments, (10())-oriented bulk silicon.
75As-doped to 1-1.8 Q -cm, were used. The n-type SOS samples were epitaxially deposited
5
Care was taken to record the emission spectra within fivc minutes of samnple pr.pairatIoM in
order to minimize the degradation of the luminescence, which in some cas,,e is ohsecrx.d %%tth
exposure to air. The porous SOS samples show photoluminescence signals comparable to
those published for p-type bulk silicon [14.15]. The photoluminescence spectra of 1(I-rm-thick
SOS etched for 9 minutes are shown in figure 3. The two curves shown arc the emission ,pctra
obtained when the sample was illuminated and emission ,as collected at the silicor ',ide
(dotted line) and at the sapphire side (solid line) of the wafer. The photoluminc,cenc_ maxi-
mizes in intensity at - 700 nm with a width of -10oo nm. The luminescence from the
Cr+ 3 impurity (695 nm) is pronounced in the sapphire-side illumination spectrum. The
similarity in the front and backside spectra suggests that there is uniformity in the porou,
structure with depth and that strain effects due to the lattice mismatch between the silicoon and
the sapphire are minimal.
I I II
SILICON SIDE
(,-" .' .9
I*
" "/
,
1 %
_ SAPPHIRE SIDE
U)
z
Z.
I I I I ,I ..
550 600 650 700 750 800 850
WAVELENGTH (nm)
Fabricating porous SOS allows for the accurate control of the depth of the porous laver.
Etching completely through the silicon layer to the silicon-sapphire interface allows control
of the porous layer thickness by controlling the thickness of the silicon. The latter is controlled
via either deposition techniques or oxidation/etch thinning processes, which are well-estab-
lished VLSI techniques. Experiments on 0.3-[rm-thick and 10-Ltm-thick SOS showed no signifi-
cant difference in the photoluminescence characteristics of peak emission wavelength or line-
width. The thicker porous layer did exhibit an increased luminescence intensity relative to the
thinner porous layer, which is attributed to the volumetric increase in the number of silicon
emitters.
Identically processed bulk porous silicon and porous SOS exhibit qualitatively different
emission when illuminated with a UV lamp. Bulk porous silicon, chemically etched as de-
scribed above, emits a vivid luminescence, whereas the I0-pm-thick porous SOS sample exhib-
its a hazy, diffuse luminescence emission. In order to understand these differences, the
6
morphology of the samples was analyzed by using SEM. l hc thickness of the porous lavcr i]
the bulk porous silicon is about 100 nm. Aith a surface roughnes.s of 10-20I nm. The poros
layer on the SOS sample is less discernible, but shows a comparable surface roughnes, as
shown in the SEM cross section in figure 4. Read"iy observable arc additional crevices and
cracks extending through the entire 10-gtm silicon lavcr on sapphire. 'Fhcse cracks ;'re attrib-
uted to preferential etching along threading dislocations and presumably cause diffuse §,tatter-
ing of the emitted photoluminescence. The dislocations arise from the thermal mismatch be-
tween silicon and the sapphire substratc which produces compressive stress in ihc silicron ,atvcr
during the deposition at high temperature and subsequent cooling. This stress is partiallk
relieved by the generation of threading dislocations. This hypothesis was examined by usin-g
Raman spectroscopy to measure the strain-induced splitting and shift in the O(IV) phonon. Tlhc
resulting spectra from thick and thin film porous SOS exhibit a peak at 522.2 + 0(.7 cmi
which is consistent with the relaxation of stress in the etched porous silicon film [131.
* LASER
a PS X10
z 5s e o
•.:, -,
500 550 600 650 700 750 800 85C 900 950
WAVELENGTH (nrmi
The emission spectra obtained for the porous silicon above were then compared to tne
emission from a commercial AIGaAs semiconductor laser (D.O. Industries GAL[A laser.
model 975-04-8) under an identical light-collection arrangement. The intensity of the semi-
conductor laser was specified as 5 mW, with ý,peak wavelength about 755 nm. The integrated
area wa calculated for the two spectra, shown in figure 5, and corrected ior geometrical fac-
tors. The integrated area for the semiconductor laser was 20 times that of our porous silicon
sample, implying an emission of 0.25 mW. From these results, an external quantum efficiency
of 5% was obtained.
Subsequent experiments were performed using a 10-mW IleCd laser (Liconix model
421 ONB) at 442 nm. Emission spectra were collected by using a 0.2"-m monochromenter (Acton
model 275, 150 grooves/mm grating) and a liquid-nitrogen-cooled CCD detector (Princeton
Instruments). This experimental arrangement allowed for a substantially improved detection
8
of low light level due to the multichannel collection ability and the loA dark current detector.
Typically, 5 accumulations of 1-second exposures were used to ohlain the emission Spcctra.
Several of the electrochemically etched porous silicon samples were measured, with peak
wavelengths ranging from 706 nm to 735 nm. All showed the characteristic broadhlmnd cmlis-
sion ( - 150 nm full width, half maximum) and had comparable relative intensiti-,. lThc laser
intensity was monitored during collection and the spectra were normalized to the maximum
laser intensity when appropriate. The undoped GaAs sample was reexamined for room tem-
perature photoluminescence. As many as 10 accumulations of 10 second exposue-"es \crc
acquired without evidence of photoemission. The MQW-AlGaAs'GaAs sample was also rein-
vestigated. The emission spectra were easily detected, with a peak about 843 nm. Figurc 6
compares the photoluminescence emission spectra of the MOW and porous silicon. The con-
mercial semiconductor laser was again used to quantify the magnitude of the phololumines-
cence. Correcting the data for geometrical aspects, we find that the integrated intensit\ of the
porous silicon is greater than 103 times that of the MQW structure. In one region of interest.
750 nm to 900 nm, the room temperature photoluminescence of porous silicon exceeds that of
the MQW sample by a factor of - 200. Of course, the MQW has a much smaller linewidth and
can be made to lase but for pumping applications the additional emission would be readily
traded for linewidth. The emitted intensity from the "brightest- porous silicon sample was cal-
culated to be 2.6 uW This corresponds to an external quantum efficiency of about ().L04",. sub-
stantially below that measured previously.
•MOW X 50 1
-. A%
5
uC x
z z.. ,*
I.-- .
500 550 600 650 700 750 800 850 900 950
WAVELENGTH (nm)
9
The shift to a shorter wavelength to photoexcite the samplesC xould not be expected to
cause the lower measurement for the quantum efticienc%. In fact, the increased photon ener'y
should photoexcite a larger population of states in the porous silicon, resulting in increased
photoluminescence. Fhe lower emission observed in this work from samples analyzed roughly
I month apart is attributed to an aging effect observed in porous silicon. It has been reported
[17,18] and has been observed here that oxidation of tle samples occurs with prolonged expo-
sure to air, which can be accelerated by illumination. This is evident by an increase in the Si-()
peak in the FTIR spectravwith time. We hypothesize that the oxidation process replaces surface
hydrogens with oxygens and subsequently quenches the radiative emission. ln1\c[,,Ivation of
this theory is ongoing.
RECENT RESULTS
Namavar et al. [ 19] of Spire Corporation report that the photolumincscence of their elec-
trochermically prepared porous silicon samples emit intensities "essentially equal to those rou-
tinely found at Spire for MOCVD-grown AI0 .3 Gal- As" and that the emission wavelength can
be red-shifted by increasing the concentration of germanium in porous SiI xGe alloys. Flec- -
troluminescence was obtained from an n-ITO/p-PS-junction diode when forward biased. Cor-
recting for series resistance, Namavar et al. project that the device would emit light with less
than 3 volts applied.
At the fall 1992 Materials Research Society (MRS) conference,* there were several
presentations of interest. Liu et al. (Stanford) reported on lithographically defined quantum
pillars fabricated in bulk Si. This group has reported light emission from these pillars when
substantial laser pumping (several W/cm ) is used. and attempts are now underway to make
contact with these nanostructures to investigate electroluminescence. Kozlowski et al. (Inst.
Solid State Tech., Germany) reported electroluminescence at 5 volts in photoelectrochemical-
ly etched porous pn-junctions. Light emission from the blue to the red region of the spectrum
was observed and could be tuned by varying the wavelength of incident light during the etching
process. Maruska et al. (Spire Corp.) reported red electroluminescence and the first monolith-
ic fabrication of a silicon light source (ITO/porous Si) and a silicon detector (ITO/Si). Peng et
al. (Rochester) reported a means to passivate the porous silicon surface to prevent the degra-
dation of photoluminescence by postfabrication etching in a nitric acid solution.
FUTURE DEVELOPMENTS
The investigation of the fundamental light emission mechanism in porous silicon and
porous SOS and the development of fabrication techniques that are reproducible and compat-
ible with microelectronic (VLSI) processing will be continued under the OCNR Independent
Research (IR) program at N,'COSC, RDT&E Division during FY93. Advances from this pro-
gram will be applicable to Jectroluminescent silicon and the further development of an
efficient light-emitting silicon source.
"Thtpresentations mentioned in this paragraph and related articles arc published in the Alicroclstaiinc Sen'i(ondwdror.r Aatclwnl
Science & Devices. Vol- 283. Material, Research Society Symposmim Proceedings, 1993.
10
CONCLUSIONS
Photoluminescent porous silicon and porous SOS have been fabricated by using chemical.
electrochemical, and photochemical techniques. An electrochemically fabricated porous sili-
con sample exhibited 5% quantum efficiency when pumped with 5 mW of 457.9-nm argon ion
laser light with a peak emission wavelength of 720 nm. This is consistent with other reports that
efficiencies between 1 and 10'4, have been observed. This room temperature performance
exceeds the early development of light emission from 111-V semiconductors and suggests that
continued research and development of the emission mechanism and optimization of the fab-
rication process can produce at least functionally equivalent results in a less expensive material
which is amenable to large-scale integration.
In 1966, the external quantum efficiency of direct-band gap GaAs light-emitting diodes was
reported greater than 17 [2201. The first diode lasei-pumped solid-state (CaF-':U 3- j laser
demonstrated in 1962 required the entire assembly to be placed in a liquid helium dewar due to
the requirements for the GaAs diode laser operation. Since then. a 100-mW AIGaAs diode
array was commercially produced in 1984, and subsequently the output power has increased
two orders of magnitude to 10 watts in multistripe bar arrays commercially available [2.31.
Extrapolation of similar rapid development can be made for porous silicon due to the mature
silicon processing infrastructure in-place worldwide. This is evident by the monolithically inte-
grated silicon-based source and detector and the rapid advances in silicon LEDs. Improve-
ments in fabrication techniques, particularly to tune the emission spectra and increase
emission, have already begun and offer significant promise as an alternative solid-state laser
pump source.
REFERENCES
11
S. R. P.Vasquez, R. WV. Fathauer.T Ge(org~e. A. Ksendzov T1.L. L.in. "Electronilc Slruciure: of
Light-Emitting Porous Si." App!. Phys. Lett.. 60 (8). F'eb). 199)2. pp. 10014- 10016,
9. N. L. Zheng, WV. Wang. H. C. Chen, 'Tomlus minperature Dependencies of Phoro(,lu-
minescence for Visible-Light -Emitting Porous Si." Appi. Phys. Leitt- 60 (2). Vebh, 1992, pp.
986-988.
10. M. S. Brandt, H. D. Fuchs, MI. Stutzmann. J. Weber. ),I.(iardo)na. "The (.ricinofVil
Luminescence From 'Porous Silicon': A Niewx Interpretation. Solid Stiate (oninun.. X1
(4), 19921, pp. 307 -3 )12.
11. N. Koshida, H. Kovama, "Visible Electrolumninescence from Porous Silicon.'"AppI. Phys.
Lett., 60 (3), Jan. 19921. pp. 347-349.
12. H. P.Maruska, F. Namavar. N. M. Kalkhoran. "Current Injection Mkechanismi for Porous-
Silicon Transparent Surface Light-Emitting Diodes." Appi. Phys. Lett_. 61 (11 ). 14 Sept.
1992, 1338- 1340.
13. W B. Dubbelday. D. M. Szaflarski, R. Shimahukuro, S. 1.. Russell. MI. J. Sailor. "Photoln-
minescent Thin Film Porous Silicon on Sapphire." Appi. Phvs. Lett.. 621 (141). Apr. 1993,
pp. 1694 -1696.
14. J. Sarathy, S. Shih, K. Jung, C. Tsai, K.-H. LI, D.-L. Kwong. J. C. Campbell. S.-L. Yau. A. J.
Bard, "Demonstration of Photoluminescence in Nonanodized Silicon," AppI. Phys. Lett.,
60 (13), Mar 1992, pp. 1532-1534.
15. R. W. Fathauer, T George, A. Ksendzov. R. P.Vasquez, "Visible Luminescence From Sili-
con Wafers Subjected to Stain Etches," Appi. Phys. Lett.. 60 (8). Feb. 1992. pp. 995 -997.
16. S. K. Ghandi, VL SI FabricationPrinciples, (New York: Wiley. 1983), pp. 478 -482.
17. R. T. Collins, M. A. Tischler, and J. H. Stathis, "Photoinduced 1Ed rogen Loss From Po-
rous Silicon," Appi. Phys. Lett., 61 (14), Oct. 1992, pp. 1649-1651I.
18. C. Tsai, K.-H. Li, J. Sarathv, S. Shih, J. C. Campbell, B. K. Hance. I. M. White. "Thermal
Treatment Studies of the Photoluminescence Intensitv of Porous Silicon,'" AppI. Phys.
Lett., 59 (22), Nov. 1991, pp. 2814-2816.
19. E Namavar, H. P.Maruska. N. M. Kalkhoran, "Visible Electroluminescence From Porous
Silicon np}-leterojunction Diodes," App!. Phys. Lett., 60 (20), May 1992. pp. 2514 -2516.
20. H. Rupprecht, J. M. Woodall, K. Konnerth, D. G. Pettit, "Efficient Electroluminescence
From GaAs Diodes at 3(00K," AppI. Phys. Lett., 9 (6) Sept. 1966, pp. 221- 223.
INITIAL DISTRIBUTION
Code 0012 Patent Counsel (1)
Code 0141 A. Gordon (1)
Code 50 H.O. Porter (1)
Code 55 H.E. Rast (1)
Code 553 G. A. Garcia (1)
Code 553 R. L. Shimabukuro (5)
Code 553 W. B. Dubbelday (5)
Code 553 S. D. Russell (15)
Code 804 V. N. Smiley (1)
Code 843 D. A. Gookin (1)
Code 961 Archive/Stock (6)
Code 964B Library (2)
oubliC reporlng burden for this collection of nformation is estimated to average I hour per esoore wu,..Or i me t e e 12' '( ,, r Ii. 1 •' 2.Ca'2 I r', .l' OaUr. ,n- ,A n- ,
ma•ntaining the data reedeo, and compieting and reviewing the collection of nformation Senr comrnaets egardlg t-wtnarn
II'S o~roen a ' ; aN-n' 2' 22 -
21 .r,,' ., ,
suggestions for reducing this bturde, to Wasfhmngtor Headquarters Ser'aces Directoraie tor Inlontcalor Operatori, arc Nftp.plci " e2t
.efv'v'' U ._ r' w. S t" .'
L, At,, ':" A
222()2-4302. and to the Ofitce of Management and Budget. Paperworkr Reductror, Pro~ect tOla 0' 8 asnlr DC 200)ý03
I1 AGENCY USE ONLY (Leave btank) 2 REPORT DATE 3 REPORT TYPE ANC OATES CO. LPE:n
IMarch 1993 Jun -- Sep 1992
4 TITLE AND SUBTITLE L F N%:!G %UMBERRS
Preliminary experiments were performed to investigate the feasibility of using silicon-based sources to pump solid-state
lasers. Photoluminescent porous silicon and porous silicon-on-sapphire (SOS) have been fabricated using chemical, electrochemi-
cal, and photochemical techniques. An electrochemically fabricated porous silicon sample exhibited 5Y quantum efficiency when
pumped with 5 mW of 457.9 nm argon ion laser light with a peak emission wavelength of 720 nm. This is consistent with other
reports that efficiencies between 1 and 10'7 have been observed. This room temperature performance exceeds the early develop-
ment of light emission from III-V semiconductors and suggests that continued research and development of the em .oi-on- rn.ecnt,
nism and optimization of the fabrication process can produce, at the very least, functionally equivalent results in a lhss expensive
material which is amenable to large-scale integration.
UNCLASSIFIED