IN THE UNITED STATES PATENT AND TRADEMARK OFFICE
`
`In re Patent of:
`
`Lebens et al.
`
`U.S. Patent No.: 6,488,390
`
`
`
`Issue Date:
`
`December 3, 2002
`
`Appl. Serial No.: 09/978,760
`
`Filing Date:
`
`October 16, 2001
`
`Title:
`
`COLOR-ADJUSTED CAMERA LIGHT AND METHOD
`
`
`
`
`
`
`
`PETITION FOR INTER PARTES REVIEW OF UNITED STATES PATENT
`
`NO. 6,488,390 PURSUANT TO 35 U.S.C. §§ 311–319, 37 C.F.R. § 42
`
`
`
`
`
`Exhibit LG-1007
`
`Appl. Phys. Lett. 71(10), 8 September 1997, American Institute of Physics
`(“Basrur”);
`
`
`
`

`

`The process and efficiency of ultraviolet generation from gallium nitride blue light
`emitting diodes
`J. P. Basrur, F. S. Choa, P.-L. Liu, J. Sipior, G. Rao, G. M. Carter, and Y. J. Chen
`
`Citation: Applied Physics Letters 71, 1385 (1997); doi: 10.1063/1.119901
`View online: http://dx.doi.org/10.1063/1.119901
`View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/71/10?ver=pdfcov
`Published by the AIP Publishing
`
`Articles you may be interested in
`3D numerical modeling of the carrier transport and radiative efficiency for InGaN/GaN light emitting diodes with
`V-shaped pits
`AIP Advances 6, 055208 (2016); 10.1063/1.4950771
`
`Mechanism of hole injection enhancement in light-emitting diodes by inserting multiple hole-reservoir layers in
`electron blocking layer
`J. Appl. Phys. 119, 105703 (2016); 10.1063/1.4942881
`
`How to decide between competing efficiency droop models for GaN-based light-emitting diodes
`Appl. Phys. Lett. 107, 031101 (2015); 10.1063/1.4927202
`
`Current crowding effect on the ideality factor and efficiency droop in blue lateral InGaN/GaN light emitting diodes
`Appl. Phys. Lett. 97, 251110 (2010); 10.1063/1.3529470
`
`Carrier dynamics in nitride-based light-emitting p-n junction diodes with two active regions emitting at different
`wavelengths
`J. Appl. Phys. 94, 2167 (2003); 10.1063/1.1591051
`
`
`
` Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 27.122.242.71 On: Fri, 07 Oct 2016
`02:35:57
`
`Exhibit LG-1007 Page 1
`
`

`

`The process and efficiency of ultraviolet generation from gallium nitride
`blue light emitting diodes
`J. P. Basrur, F. S. Choa,a) P.-L. Liu,b) J. Sipior,c) G. Rao,d) G. M. Carter, and Y. J. Chen
`Department of Computer Science and Electrical Engineering, University of Maryland, Baltimore
`County, Baltimore, Maryland 21250
`共Received 3 January 1997; accepted for publication 8 July 1997兲
`To obtain small size, high speed ultraviolet sources, we studied the ultraviolet 共UV兲 generation
`process and efficiency of gallium nitride 共GaN兲 blue light emitting diodes 共LEDs兲. The blue and UV
`emissions follow a four-level recombination model. Depending on the pump pulse amplitude, the
`UV-to-blue generation ratio increases and then saturates with increasing pump pulse duration. High
`efficiency, up to 450 ␮W UV power at 380 nm, can be obtained from a 1.2 mW blue LED. © 1997
`American Institute of Physics. 关S0003-6951共97兲04736-0兴
`
`The quest for compact ultraviolet 共UV兲 light sources has
`been driven by their potential use in portable fluorescence
`intensity and lifetime sensors, drug testing,1 and other medi-
`cal applications.2 Various materials based on GaN are cur-
`rently being investigated for optical devices operating in the
`UV and blue-violet region. Intensity dependent photolumi-
`nescence in GaN thin films,3 and time resolved spectra of
`GaN light emitting diodes 共LEDs兲 have been studied
`recently.4–6 The blue LEDs with either an InGaN double
`heterostructure active layer 共Nichia, PA兲 or a Zn disordered
`GaN active layer 共CREE, NC兲 normally emit blue light. It
`was understood that
`the blue emission is generated by
`electron-hole recombination from the doping level to the
`ground level. When pumped with short electrical pulses, UV
`emission was also generated through the band edge recom-
`bination. The UV generation efficiency strongly depended on
`the characteristics of the electrical pump pulses. Since, cur-
`rently, there are no undoped UV LEDs commercially avail-
`able, we have to use blue LEDs as our UV sources. To un-
`derstand the generation process and optimize generation
`efficiency of UV light with any given electrical pump source,
`we performed both theoretical and experimental studies of
`the generation process.
`The LEDs we used are the Nichia InGaN, double hetero-
`structure, blue LED with 50-␮m-thick Zn doped InGAN ac-
`tive layer sandwiched by two AlGaN layers.7 The output
`spectra of the blue LEDs were measured under both dc bi-
`ased and pulsed operation with a SPEX 1250M monochro-
`mator and a photomultiplier tube detector. The LEDs were
`soldered to SMA connectors and driven through a bias T
`with a dc current source and Hewlett-Packard 214B pulse
`generator capable of generating voltage pulses up to 50 V
`with 10 ns minimum pulse duration. The Nichia blue LED
`was found to produce very weak UV light when the dc bias
`was increased above 30 mA. More intense UV light could be
`generated from the LED under pulsed operation. The UV-to-
`
`a兲Electronic mail: choa@umbc.edu
`b兲On leave at Naval Research Laboratory, Washington DC 20375 from De-
`partment of Electrical and Computer Engineering, SUNY at Buffalo, Am-
`herst, NY 14260.
`c兲Department of Chemical and Biochemical Engineering, UMBC.
`d兲Department of Chemical and Biochemical Engineering, UMBC and the
`Medical Biotechnology Center of the Maryland Biotechnology Institute,
`University of Maryland at Baltimore, Baltimore, MD 20201.
`
`blue ratio varied with the amplitude and duration of the
`pump pulse. Under dc operation, the LED was destroyed
`before significant levels of UV light were generated.
`Figure 1 shows the output spectra of the Nichia LED
`generated with 10 ns pump pulses at various amplitudes. The
`injected carriers are trapped by the upper states of the blue
`band due to a very short transition time from the band edge
`to these states. When these states are nearly filled up, the UV
`recombination becomes more effective. By increasing the
`pump pulse amplitude more carriers are provided for the UV
`generation. Figure 2 shows the output spectra generated with
`50 V pulses at various pulse durations. The carrier injection
`rate remains unchanged since the pump pulse amplitude is
`constant. However, after filling most of the upper states of
`the blue band, the remainder of the pump pulse generates
`carriers available for UV light generation. Therefore, to ob-
`tain more UV light, either the pump pulse amplitude or the
`duration can be increased.
`Both simulation and transient studies have been done to
`understand the generation and recombination mechanism and
`estimate the recombination parameters. Figures 3共a兲 and 3共b兲
`show, respectively, the optical pulse shapes at 380 and 450
`nm when the LED is pumped with 50 V, 100 ns pulses
`without dc bias. Time resolved spectral studies and optical
`outputs of shorter electrical pump pulses can be found in our
`previous work.6
`The UV and blue generation mechanism can be de-
`scribed by the four-level system shown in Fig. 4. Both the
`N 1 and N 2 levels 共equal in number兲 are created by Zn dop-
`ants as the usual acceptor states of Zn and shallow states near
`the conduction band. The N 2 states have to be included to
`explain the UV and blue generation behavior. Using a three-
`level system, which lacked the N 2 states, we found it is im-
`possible to fit the time response to the experimental results,
`as shown in Figs. 3共a兲 and 3共b兲.
`The four-level recombination process can be described
`below:
`
`dN 3共t 兲
`dt
`
`⫽
`
`J共t 兲
`qd
`
`⫺
`
`N 3共t兲
`␶30
`
`⫺
`
`N 3共t 兲
`␶32
`
`⫺
`
`N 3共t 兲
`␶non3
`
`,
`
`dN 2共t 兲
`dt
`
`⫽
`
`N 3共t 兲
`␶32
`
`⫺
`
`N 2共t 兲
`␶21
`
`,
`
`共1兲
`
`共2兲
`
`Appl. Phys. Lett. 71 (10), 8 September 1997
`
`0003-6951/97/71(10)/1385/3/$10.00
`
`© 1997 American Institute of Physics
`
`1385
`
` Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 27.122.242.71 On: Fri, 07 Oct 2016
`02:35:57
`
`Exhibit LG-1007 Page 2
`
`

`

`FIG. 3. 共a兲 Sampling trace of the emission at 380 nm, 共b兲 sampling trace of
`the emission at 450 nm.
`
`FIG. 1. Output spectrum of blue LED at 40 ns for various pulse amplitudes.
`
`⫽
`
`⫺
`
`,
`
`共3兲
`
`N 1共t 兲
`N 2共t 兲
`dN 1共t兲
`␶10
`␶21
`dt
`where N i is the population of the ith level, i⫽0,1,2,3, J is
`the current density, d is the thickness of the active layer, ␶i j
`represents the lifetime of the transition from level i to level
`j, and ␶non3 represents the nonradiative recombination life-
`time at band edge. While the lifetimes ␶30 and ␶10 are as-
`sumed to be constant, the lifetimes ␶32 and ␶21 are dependent
`on N 2 and N 1 , respectively. More specifically they can be
`expressed in terms of the empty band recombination lifetime
`
`⬘ and ␶⬘ as3
`2
`21
`
`␶3
`
`N max
`N max⫺N 2共t 兲
`N max
`N max⫺N 1共t 兲
`where N max is the dopant concentration. From the above two
`equations we can infer that ␶32 and ␶21 increase as their
`
`⬘ ,
`2
`
`␶3
`
`⬘ ,
`1
`
`␶2
`
`␶32⫽
`
`␶21⫽
`
`共4兲
`
`共5兲
`
`respective bands are filled. The transition will become infi-
`nitely slow when N 1 or N 2 approaches N max . To simplify the
`model we have neglected the higher order nonradiative car-
`recombinations 共Auger and biomolecular兲. Simulta-
`rier
`neously solving Eqs. 共1兲, 共2兲, and 共3兲 we obtain the time
`dependent population of N 3 , N 2 , and N 1 for any given input
`injection current density J(t). The instantaneous UV and
`blue light intensities are given by3
`N 2共t兲
`␶21
`
`I blue共t 兲␣
`
`hc
`␭ blue
`
`,
`
`共6兲
`
`FIG. 2. Output spectrum of blue LED at 50 V for various pulse durations.
`
`FIG. 4. Four-level recombination model of the GaN blue LED.
`
`1386
`
`Appl. Phys. Lett., Vol. 71, No. 10, 8 September 1997
`
`Basrur et al.
`
` Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 27.122.242.71 On: Fri, 07 Oct 2016
`02:35:57
`
`Exhibit LG-1007 Page 3
`
`

`

`FIG. 6. The UV to blue ratio as a function of pulse duration and amplitude.
`
`With a Nichia LED rated at 1.2 mW blue light output,
`we have obtained 450 uW of UV light output when it is
`pumped with 50 ns, 50 V pulses at a 2.5 MHz repetition rate.
`No damage to the LED was apparent. The center wavelength
`of the UV band shifted from 380 to 395 nm under such
`operation and the measured UV power was limited by this
`thermal shifting due to the increased attenuation of the opti-
`cal notch filter used in the measurements.
`A very important factor determining the UV-to-blue ra-
`tio is N max . Provided no other recombination mechanisms
`come into play, reducing the doping should increase the UV
`light generation. It would therefore be possible to produce a
`UV LED that has little blue light production by simply re-
`ducing the doping density during manufacture of the LED.
`Currently no UV LED based on undoped GaN is available
`commercially.
`In conclusion, to obtain compact size UV sources, we
`have studied the UV generation efficiency from blue GaN
`LEDs. Short electrical pulses facilitate the generation of UV
`light. High efficiency and high power up to 0.45 mW UV has
`been generated from a 1.2 mW blue LED. A four-level re-
`combination model is presented to explain the process. Using
`the four-level model we can predict the UV generation effi-
`ciency with different pump pulse characteristics.
`G. Rao acknowledges support from NSF Grant No.
`BES-9413262 and NIH Grant No. RR10955.
`
`FIG. 5. Theoretically computed time response of the UV and blue light
`outputs.
`
`I UV共t 兲␣
`
`hc
`␭ UV
`
`N 3共t兲
`␶31
`
`.
`
`共7兲
`
`Using the above equations, the blue and UV pulses are plot-
`ted in Fig. 5. By setting the values of the parameters ␶30 ,
`
`⬘ , ␶⬘ , ␶10 , and the ␶non3 along with the doping concen-
`2
`21
`tration N max , a best fit to the experimental data was obtained,
`as shown in Table I. The value of ␶30 was set to 5 ns which
`was derived from our previous time resolved spectral
`studies.6 The value of N max was estimated from the UV gen-
`eration threshold of 30 mA.
`Based on the fitted parameters, we can calculate the UV-
`to-blue ratio as a function of the duration and amplitude of
`the pump pulse. Since the LED has a nonlinear I – V curve,
`the input impedance, and thus the current, will depend upon
`the pulse voltage. Under 50 V pulse operation, the LED has
`an input impedance of 67 ⍀. Figure 6 gives the production
`efficiency of UV light as a function of the pump pulse am-
`plitude and duration. The UV-to-blue light ratio increases
`with both the pump pulse amplitude and duration. Using
`pump pulses with larger amplitude has the advantage of gen-
`erating UV with higher efficiency. In addition, the slope of
`the UV-to-blue ratio versus duration is high for short pulses
`and then saturates as the duration increases. Hence the im-
`provement in efficiency obtained from increasing the pump
`pulse duration decreases. Since the diode will typically burn
`out at 120 mA dc, the pump pulse duration cannot be in-
`creased indefinitely. Thermal effects also become apparent
`with long pump pulse durations, decreasing light generation
`efficiency and redshifting the emission.
`
`␶3
`
`Value
`
`5 ms
`0.1 ns
`115.5 ns
`0.821 ns
`55 ns
`6⫻1018
`
`1 F.-S. Choa, M.-H. Shih, A. R. Toppozada, M. Block, and M. E. Eldefrawi,
`Anal. Lett. 29, 29 共1996兲.
`2 J. Sipior, L. Randers-Eichhorn, J. R. Lakowicz, and G. Rao, Biotechnol.
`Prog. 12, 266 共1996兲.
`3 R. Singh, R. J. Molnar, M. S. Unlu, and T. D. Moustakas, Appl. Phys.
`Lett. 64, 3 共1994兲.
`4 T. Araki and H. Misawa, Rev. Sci. Instrum. 66, 12 共1995兲.
`5 G. Mohs, B. Fluegel, H. Giessen, H. Tajalli, N. Peyghambrian, P.-C. Chiu,
`B. S. Phillips, and M. Osinski, Appl. Phys. Lett. 67, 1515 共1995兲.
`6 F. S. Choa, J. Y. Fan, P.-L. Liu, J. Sipior, G. Rao, G. M. Carter, and Y. J.
`Chen, Appl. Phys. Lett. 69, 3668 共1996兲.
`7 S. Nakamura, T. Mukai, and M. Senoh, Appl. Phys. Lett. 64, 1687 共1994兲.
`
`TABLE I. Parameter value.
`
`Parameter
`
`␶30
`⬘
`2
`⬘
`1
`␶10
`␶non3
`N max
`
`␶2
`
`␶3
`
`Appl. Phys. Lett., Vol. 71, No. 10, 8 September 1997
`
`Basrur et al.
`
`1387
`
` Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 27.122.242.71 On: Fri, 07 Oct 2016
`02:35:57
`
`Exhibit LG-1007 Page 4
`
`