Litton C.W., . Reynolds D.C., Collins T.C. Zinc Oxide Materials for Electronic and Optoelectronic Device Applications

Подождите немного. Документ загружается.

At room temperature laser emission at 2.75 eV was found at a threshold of 350 kW cm

2

for the same sample. Other samples with y ¼0.16 and well width 2.7 nm were found to lase

with emission peaks as low as 2.53 eV corresponding to emission wavelengths around

490 nm [Figure 10.4]. The lasing thresholds are found to be relatively robust to tempera-

ture increas es, a feature indicative of gain mechanisms related to the recombination of

carriers localized within the quantum wells. These are encouraging developments for a

ZnO-based technology, with emission extending well into the green, there is pote ntial

advantage over GaN-based systems, though this work is at an early stage.

(Zn,CD)O/ZnO, MQWs (xn), T = 5 K

350

300

250

200

150

100

0 50 100 150

T [K]

Photon energy [eV]

PL intensity [arb. units] PL intensity [arb. units]

Lasing threshold [kW/cm

]

200 250 300

(a)

(b)

2.3 2.4 2.5 2.6 2.7 2.8 2.9

(5)

exc

= 1 MW/cm

0.10

0.15

0.17

0.16

2.6 nm, (x7)

3.2 nm, (x7)

2.4 nm, (x5)

2.0 nm, (x10)

2.7 nm (x10)

(xn)

(4) (3) (2)

290 K 5K

(1)

Figure 10.4 Laser emission spectra from ZnCdO MQWs. (a) Variation of the laser emission

with structure design; 1–4 are on sapphire substrates, 5 is on ZnO substrate. (b) Change of laser

emission from low to room temperature for structure 2. Inset: temperature dependence of the

threshold.

[46]

Stimulated Emission 273

10.4 Zinc Oxide Lasers

10.4.1 Introduction

According to modern convention demonstrations of “laser action” or “lasing” requires

coherence, and this is conventionally achieved by placing gain media inside an optical

cavity. This cavity will then favor the growth of particular frequencies and phases. Thus

selective ampliﬁcation is a result of positive feedback for electromagnetic waves that form

a standing pattern in the cavity. In gener al, to prove las ing it is necessary to provide

evidence of the formation of Fabry–Perot modes.

Mode spacing (Dl) is given by:

Dl ¼

2Lnl

dn=dlðÞ

where L is the cavity length and n is the index of refraction at l

(the wavelength of one of

the modes).

Meanwhile, to a ﬁrst approximation, the gain threshold (G

) is given by:

2bL

where a is the loss per unit length, b is the gain and R

and R

are the mirror (or facet)

reﬂectivities.

To observe modes directly it is desirable to have a short cavity length as this will

increase the mode separation but small cavity lengths require higher threshold intensities.

High mirror reﬂectivities are very desirable as this reduces thresholds. Resolution of laser

modes from cavities prepared by simple cleaving is difﬁcult, as the process tends to

produce poor quality mirrors with low reﬂectivities, and relatively large cavity lengths and

this combined with excitation pulse variations, and spatial variations in the sample often

produces broad Gaussian and Lorentzian lineshapes when laser spectra are generated by

boxcar integration or multiple charge-coupled device (CCD) acquisitions. To obtain

spectra with clear mode structure it is necessary to resolve emissions spatially and

temporally as well as spectrally.

[47]

Thus, excitation stripes should be narrow to

avoid the effect of spatial variations along the laser bar and the emission from a single

excitation pulse should be collected and dispersed into a high resolution monochromator/

CCD system.

The ﬁrst room temperature ZnO laser was reported in 1997 by the Tohoku University

group.

[20]

Samples were grown by plasma-enhanced MBE on sapphire substrates. To

produce laser cavities as-grown samples are cleaved to produce “bars” with cavity lengths

in the range 300–1000 mm. By moving an intense excitation stripe, generated by a

frequency tripled Nd:YAG laser, along the bar high intensity regions emitting speckled

light were located. When these regions were studied in detail a threshold for stimulated

emission was determined at 240 kW cm

2

and lasing was proven by the presence of

modulations in the spectra taken from single excitation pulses. As far as we are aware there

is only one other instance where a ZnO cavity has been fabricated; however, one of the

274 Room Temperature Stimulated Emission and ZnO-Based Lasers

most fascinating features of ZnO is its almost unique tendency to form natural laser

cavities. “Microstructural”, “powder” and “nanowire” ZnO lasers have all been reported

within the last few years.

10.4.2 Microstructural Lasers

There have now been many reports of ZnO laser cavities formed as a result of

microstructural features of thin ﬁlms. The Tokyo Institute of technology and their

collaborators

[19,22,28,29]

found regularly spaced sharp lines in lasing spectra reminiscent

of cavity modes (Figure 10.5.). This was surprising since there had been no attempt to form

a las er cavity. Furthermore, the mode spacing and therefore cavity length was found to

1.0

0.8

0.6

0.4

0.2

0.0

I=1.121

I=1.021

He-Cd

295K

70K

0.0

3.0 3.1

Photon Ener

(eV)

3.2

Emission Intensity (a.u.)

3.3 3.4 3.5

1.0

F ⊥c

–I

F//c

)(I

B_C

E//c

)

2.0

I/I

Figure 10.5 The lower trace shows the absorption spectrum (dotted curve) and photolumi-

nescence spectrum, measured at 70 K and room temperature, respectively. The upper traces

show spontaneous and stimulated emission spectra under pumping intensities of 0.2I

, 1.02I

and 1.12I

provided by the frequency-tripled and mode-locked Nd:YAG laser (355 nm, 15 ps).

The inset on the right-hand side shows the degree of polarization of the UV emissions plotted as

a function of the pumping intensity. Reprinted from Z. K. Tang, et al., Room-temperature

ultraviolet laser emission from self-assembled ZnO microcrystallite thin ﬁlms, Appl. Phys. Lett.

72, 3270. Copyright (1998) with permission from American Institute of Physics

Zinc Oxide Lasers 275

change, in steps, with the length of the excitation stripe (Figure 10.6.). The authors

concluded that the laser cavity was formed by the {1100} facets of parallel hexagonal ZnO

microcrystallites, where many hexagons are covered by the excitation stripe. The

necessary mirrors were described as a result of the combination of the change in refractive

index due to excitation and the grain boundaries.

Workers at Northwestern University have reported laser action from nonepitaxial ZnO

(Figure 10.7).

[48]

Polycrystalline ﬁlms were grown on amorphous fused silica substrates by

laser ablation. They found laser cavities to “self-form” as a result of strong optical

scattering in the ﬁlms. The scattering mean free path is estimated to be of the order of the

emission wavelength of ZnO and it is suggested that emitted light could return via a

“random” or complex path to the scatterer from which it was scattered and thus form a

3.6

2.4

1.2

0.0

–90 –60 –30 0 30 60 90 120

I/L (mm

–1

)

Angle (degree)

Laser Emission

Laser Emission Intensity (a.u.) Mode Spacing (meV)

Pumping

(a)

(b)

10 15

Sample

Figure 10.6 (a) The mode spacing (solid circles) plotted as a function of the stripe length L of

the pumping laser beam. The solid line is calculated for a Fabry–P



erot resonant cavity with

length L. A model of natural Fabry–P



erot microcavity formed by the hexagonal microcrystallites

is schematically drawn in the inset. (b) The laser emission intensity plotted as a function of

rotating angle. The experimental set-up is shown in the inset. The maximum and minimum of

the laser emission intensity repeats every 60



. Reprinted from Z. K. Tang, et al., Room-

temperature ultraviolet laser emission from self-assembled ZnO microcrystallite thin ﬁlms, Appl.

Phys. Lett. 72, 3270. Copyright (1998) with permission from American Institute of Physics

276 Room Temperature Stimulated Emission and ZnO-Based Lasers

closed-loop path (Figure 10.7, inset). A number of subsequent publications by North-

western and collaborators have examined

[49–51]

and modelled

[52]

random cavity formation.

Models are suggested in which regions of net gain, regions of loss and scattering centers

can lead to the formation of cavities.

Other groups have also reported lasing as a result of microstructural cavities. Mitra and

Thareja

[53]

have observed lasing from ZnO pellets and polycrystalline thin ﬁlms formed by

pulsed laser deposition. Cho et al.

[54]

have observed lasing from polycrystalline ﬁlms

deposited by the oxidation of metallic zinc.

10000

8000

6000

4000

2000

3000

2000

1000

380 382 384 386

Wavelen

th (nm)

(a)

(b)

(c)

Emission Intensity (a.u.)Emission Intensity (a.u.)

388 390 392 394 396

Figure 10.7 Emission spectra from a ZnO “powder laser” when the excitation area is (a)

2700 mm

, (b) 3800 mm

and (c) 4500 mm

. The excitation intensity is 400 kW cm

2

. The inset is

a schematic diagram showing the formation of a closed-loop path for light through multiple

optical scattering in a random medium. Reprinted from H. Cao, Y. G. Zhao, H. C. Ong, S. T.

Ho, J. Y. Dai, J. Y. Wu, and R. P. H. Chang, Ultraviolet lasing in resonators formed by scattering

in semiconductor polycrystalline ﬁlms, Appl. Phys. Lett., 73, 3656. Copyright (1998) with

permission from American Institute of Physics

Zinc Oxide Lasers 277

What is clear from all these studies is that grain boundaries, features normally avoided,

in ZnO can play a signiﬁcant role in the formation of las er cavities. Spatially resolved

electron energy loss spectroscopy (EELS) has revealed signiﬁcant changes in refractive

index to form reﬂective layers of around 10 nm in the vicinity of grain boundar ies.

[55]

This change is speculated to be due to the formation of space-charge regions at the grain

boundaries.

10.4.3 Powder Lasers

ZnO has also provided the ﬁrst demonstration of laser action in a powder.

[56]

ZnO

nanoparticles were produced by the precipitation reaction of zinc acetate dihydride and

diethylene glycol. The nanocrystals of around 50 nm agglomerate to form micrometer-

sized clusters (Figure 10.8). Excitation with the 4th harmonic of a Nd:YAG laser (266 nm)

above a threshold of around 0. 3 nJ at room temperature results in a very sharp peak

(FWHM ¼12 nm) in the emission spectrum at around 380 nm (Figure 10.9). Similar to the

lasing observed in polycrystalline ZnO ﬁlms, optical cavities are formed by multiple

scattering and wave interference within the disordered medium. Whether these lasers

are simply complex cavities or whether the Anderson localization plays a part is still a

matter for debate.

Figure 10.8 Spectrally integrated intensity of emission from the ZnO cluster vs the incident

pump pulse energy. The inset is the SEM image of the ZnO cluster. Reprinted from H. Cao, J. Y.

Xu, E.W. Seelig, and R.P.H. Chang, Microlaser made of disordered media, Appl. Phys. Lett., 76,

2997. Copyright (2000) with permission from American Institute of Physics

278 Room Temperature Stimulated Emission and ZnO-Based Lasers

10.4.4 Nanowire Lasers

Perhaps one of the most intriguing developments over the last few years has been the

demonstration of room temperature ZnO nanowire “nanolasers”.

[57,58]

Physical vapor

deposition techniques can be used to fabricate arrays of ZnO nanowires via the vapor–

liquid–solid (VLS) process (Figure 10.10).

[57]

Sapphire substrates are coated with a thin

Figure 10.9 (a, c, e) Emission spectra from the ZnO cluster shown in Figure10.8. (b, d, f) The

corresponding spatial distributions of emission intensity in the cluster. The incident pump pulse

energy is 0.26 nJ for (a) and (b), 0.35 nJ for (c) and (d), and 0.50 nJ for (e) and (f). Reprinted

from H. Cao, J. Y. Xu, E.W. Seelig, and R.P.H. Chang, Microlaser made of disordered media,

Appl. Phys. Lett., 76, 2997. Copyright (2000) with permission from American Institute of

Physics

Zinc Oxide Lasers 279

ﬁlm of gold, as the substrate reaches the growth temperature the gold forms nanoclusters

and subsequent ZnO wire growth is catalyzed by the gold. Very well faceted hexagonal

wires with excellent crystal quality are commonly produced. The distribution of wire

diameters, from 20 to 150 nm, reﬂects the distribution of gold clusters, though 95% of

wires have diameters between 70 nm and 100 nm. Lengths are readily controlled up to

10 mm. When optically excited by the frequency quadrupled output of a Nd:YAG laser

(266 nm) a threshold of 40 kW cm

2

and clear mode structure in emission spectra indicate

lasing due to exciton–exciton scattering [Figure 10.11(a,b)]. The length of the nanowire

forms the laser cavity length and the sapphire/ZnO and ZnO/air interfaces at either end of

the wires provide the mirrors [Figure 10.11(c)].

The low threshold for laser action in ZnO can be attributed to the single mode nature of

the lasers (fortuitously wires with diameters in the range 80–120 nm are single mode

waveguides for UV light), the giant oscillator strength effect and the high crystal quality

(which also manifests itself in long lumin escence lifetimes). A lot of nanotechnological

ingenuity will be required to turn these nanolasers into practicable devices but the

prospects for integration into “system-on-a-chip”, use as optical probes or pixelated

displays are enticing. In addition to offering a glimpse of a nanotechnological future the

work of the Berkeley group

[57]

also independently veriﬁes that stimulated emiss ion and

lasing can be achieved in ZnO crystals with thresholds as low as 40 kW cm

2

[18]

10.4.5 ZnO Laser Diodes

Both the fabrication of p-type ZnO and the reproducibility of laser-quality ZnO remain

seriously challenging and, as yet, only one group has reported on the fabrication of ZnO-

based laser diodes.

Figure 10.10 (a–e) SEM images of ZnO nanowire arrays grown on sapphire substrates. A top

view of the well-faceted hexagonal nanowire tips is shown in (e). (f) High-resolution TEM image

of an individual ZnO nanowire showing its <0001> growth direction. Reprinted from M.

Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo, P. Yang, Science, 292,

1897. Copyright (2001) with permission from AAAS

280 Room Temperature Stimulated Emission and ZnO-Based Lasers

Ryu et al.

[59]

fabricated arsenic-doped p-type ZnO/BeZnO and gallium-doped n-type

BeZnO/ZnO heterojunction structures with undoped Zn O/BeZnO MQW active layers.

This work represents a signiﬁcant milestone with stimulated emission reported with a

threshold of 420 A cm

2

and emission at 3.21 eV. The authors suggest exceptionally high

exciton binding energies (263 meV) in the MQ W active layers, exciton–exciton scattering

as the main emission mechanism and the existence of Fabry–Perot mode s. However,

Klingshirn et al.

[12]

have suggested alternative explanations for the experimental evidence

presented.

It is hoped that in the near future the fabrication of more devices with more detailed

analysis will yield a greater understanding and a concerted effort toward reproduci ble,

efﬁcient and lasting ZnO-based lasers, though it is clear that this is a challenging goal.

10.5 Conclusions

Although ZnO epitaxial growth technologies are still relatively immature the best reported

room temperature “ﬁgures of merit” for ZnO-based structures, 11 kW cm

2

thresholds,

300 cm

1

gain and 87 K temperature coefﬁcients, compare favorably with those of ZnSe-

Figure 10.11 (a) Emission spectra from nanowire arrays below (A) and above (B and inset)

the lasing threshold. The pump power for these spectra is 20, 100 and 150 kW cm

2

respectively. The spectra are offset for easy comparison. (b) Integrated emission intensity

from nanowires as a function of optical pumping energy intensity. (c) Schematic illustration of a

nanowire as a resonance cavity with two naturally faceted hexagonal end faces acting as

reﬂecting mirrors. Reprinted from M. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E.

Weber, R. Russo, P. Yang, Science, 292, 1897. Copyright (2001) with permission from AAAS

Conclusions 281

and GaN-based structures under similar conditions. In the context of emission character-

istics there is no fundamental reason why ZnO-based devices should not provide a

competitive alternative to GaN-based devices in some if not all circumstances. The main

issues remain reproducibility of “laser quality” and p-type material as part of complete

laser structures. There is still plenty of room for progress in this regard as the total research

investment in ZnO laser technology is still a tiny fraction of the total investment put in to

ZnSe and GaN technologies. The prevalence of the exciton–exciton gain mechanism at

room temperature, the relatively low growth temperatures, the availability of ZnO

substrates, excellent transport properties and the possibility of micro-laser self-formation

are all factors that provide justiﬁcation for continued work. However, the ZnMgCdO

system does not seem as ﬂexible in terms of band-engineering as the GaAlInN system and

issues such as device lifetimes and stability have yet to be addressed.

In this chapter we have only considered the stimulated emission and lasing with

relevance to traditional band-to-band short-wavelength devices. Two other types of device

deserve brief consideration. As yet, no experimental or theoretical work has been

published considering the potential of ZnO for use in infrared quantum cascade lasers.

With current solubility limits the band offsets in the ZnO/ZnMgO system may not allow

for devices at the communications wavelengths but with high electron densities and

excellent transport characteristics ZnO ought to have high power applications in the

2–5 mm wavelength range. A second type of device has recently received theoretical

attention. Zamﬁrescu et al.

[60]

have suggested ZnO as the semiconductor most suitable for

the realization of room temperature “polariton lasers”. The polariton laser is a device

based on the strong coupling of light with excitons in a semiconductor microcavity. This

revolutionary laser requires no population inversion to achieve optical ampliﬁcation,

instead, coherent light ampliﬁcation is based on the mechanism of the Bose condensation

of exciton polaritons. Zamﬁrescu et al. found record values for longitudinal transverse

splitting of the exciton resonances in ZnO, values two orders of magnitude larger than

those of GaAs. A ZnO-based microcavity structure is proposed that would have high gain

and thresholds as low as 3 kW cm

2

at room temperature.

References

[1] C. Klingshirn, Semiconductor Optics, Springer-Verlag, Heidelberg, 1997.

[2] J.M. Hvam, Solid State Commun. 12, 95 (1973).

[3] J.M. Hvam, Phys. Status Solid B 63, 511 (1974).

[4] C. Klingshirn, E. Ostertag and R. Levy, Solid State Commun. 15, 883 (1974).

[5] C. Klingshirn, Phys. Status Solid B 71, 547 (1975).

[6] J.M. Hvam, Solid State Commun. 26, 987 (1978).

[7] S.W. Koch, H. Haug, G. Schmieder, W. Bohnert and C. Klingshirn, Phys. Status Solid B 89, 431

(1978).

[8] J.M. Hvam, Phys. Status Solid B 93, 581 (1979).

[9] J.M. Hvam, G. Blattner, M. Reuscher and K. Klingshirn, Phys. Status Solid B 118, 179 (1983).

[10] C. Klinshirn, J. Fallert, O. Gogolin, M. Wissinger, R. Hauschild, M. Hauser, H. Kalt and H.

Zhou, J. Lumin. 128, 792 (2008).

[11] C. Klinshirn, R. Hauschild, J. Fallert and H. Kalt, Phys. Rev. B 75, 115203 (2007).

[12] C. Klinshirn, J. Fallert, H. Zhou and H. Kalt, Appl. Phys. Lett. 91, 126101 (2007).

[13] F.H. Nicoll, Appl. Phys. Lett. 9, 13 (1966).

282 Room Temperature Stimulated Emission and ZnO-Based Lasers