Ultraviolet LED (AlGaN Deep-UV LED)

1. Theory

Definition: An ultraviolet LED uses AlGaN quantum wells, usually on AlN or sapphire, to emit from the UV-A down into the deep-UV (UV-C) range.
Alloy bandgap: Raising the aluminium fraction widens the gap from GaN (3.43 eV) toward AlN ( $\approx 6.2$ eV); the bowing parameter is uncertain (about 0.7 to 1.33 eV), shifting deep-UV wavelengths by up to about 10 nm for the same composition.
Deep Mg acceptor: The acceptor deepens from about 160 meV in GaN to about 0.5 eV in AlN, so hole density falls by orders of magnitude.
Valence-band reordering: The crystal-field splitting turns negative at high aluminium content, so the top valence band becomes z-like and emission switches from TE to TM polarization.
Defects: AlGaN on sapphire has high threading-dislocation density, lowering radiative efficiency.

2. Working Principle

Emission: Carriers injected into AlGaN wells recombine at UV wavelengths set by composition, strain and confinement.
Hole injection: Poor p-doping is worked around with p-GaN contact layers, polarization-doped or superlattice layers, and tunnel junctions.
Absorption losses: The p-GaN contact layer absorbs UV light emitted toward it.
Polarization and extraction: TM-polarized light travels mostly sideways and is hard to extract through the top surface.
Low efficiency: Injection, radiative and extraction losses combine, so UV-C LEDs reach external efficiencies of only a few to about 10%, far below blue LEDs.

3. Applications

UV-C disinfection: Compact, mercury-free, instant-on sterilization of water, air and surfaces at point of use.
UV-A curing: Curing of inks, coatings, adhesives and resins in printing and manufacturing.
Sensing and analysis: UV spectroscopy, fluorescence detection and counterfeit or document checking.
Phototherapy: Narrow-band UV-B treatment of skin conditions.

Name	Formula	Related Formulae	Variables/Constants
Bandgap and emission wavelength of the AlGaN active layer	$E_{g} (A l_{x} G a_{1 - x} N) = x E_{g} (A l N) + (1 - x) E_{g} (G a N) - b x (1 - x)$	$λ (nm) = \frac{1239.8}{E ( e V )}$ $E_{e m} \approx E_{g} + E_{e 1} + E_{h 1} - E_{x}$	$E_{g} (A l_{x} G a_{1 - x} N) : ban d g a p o f t h e a l l oy, w i t h ab o u t 6.2 e V f or A l N an d 3.43 e V f or G a N (t h e b o w in g p a r am e t er b i s r e p or t e d b e tw ee n ab o u t 0.7 e V an d 1.3 e V, an d t h e c h o i ce c han g es t h e w a v e l e n g t h b y ab o u t 10 nm)$ $x : a l u mini u m f r a c t i o n$ $b : b o w in g p a r am e t er$ $λ : w a v e l e n g t h o f t h e p h o t o n, in nm f or t h e e n er g y in e V$ $E_{e m} : e mi ss i o n e n er g y o f a q u an t u m w e l l w i t h t h e co n f in e m e n t e n er g i es E_{e 1} an d E_{h 1} an d t h e e x c i t o n bin d in g e n er g y E_{x} (t h e s t r ain an d t h e p o l a r i z a t i o n f i e l d o f t h e w e l l s hi f t i t f u r t h er, a s o n t h e Q u an t u m W e l l L E D p a g e)$ $(t h e u l t r a v i o l e t ban d s a r e U V - A f r o m 315 t o 400 nm, U V - B f r o m 280 t o 315 nm an d U V - C f r o m 100 t o 280 nm, t h e I S O 21348 an d C I E co n v e n t i o n)$
Valence band ordering and polarization of the emission	$E_{1} = Δ_{c f} + \frac{Δ _{so}}{3}$	$E_{2, 3} = \frac{Δ _{c f} - Δ _{so} /3}{2} \pm (\frac{Δ _{c f} - Δ _{so} /3}{2})^{2} + 2 (\frac{Δ _{so}}{3})^{2}$ $D O P = \frac{I _{T E} - I _{T M}}{I _{T E} + I _{T M}}$ $η_{T E} = \frac{1 + D O P}{2}$	$E_{1}, E_{2}, E_{3} : e n er g i es o f t h e t h r ee v a l e n ce ban d s a t t h e z o n e ce n t r e r e l a t i v e t o t h e v a l e n ce ban d e d g e w i t h o u t s t r ain (in t h e q u a s i - c u bi c m o d e l), E_{1} ha s t h e x - y c ha r a c t er, w hi c h e mi t s l i g h t p o l a r i z e d p er p e n d i c u l a r t o t h e c - a x i s (T E), an d t h e u pp er r oo t E_{2} ha s t h e x - y c ha r a c t er f or a p os i t i v e Δ_{c f} b u t t h e z c ha r a c t er f or a n e g a t i v e Δ_{c f}, an d a z - l ik e ban d e mi t s l i g h t p o l a r i z e d a l o n g t h e c - a x i s (T M) (t h e t o p m os t ban d i s t h e n t h e z - l ik e o n e)$ $Δ_{c f} : cr y s t a l - f i e l d s pl i tt in g, s ma l l an d p os i t i v e in G a N (t e n s o f m e V or l ess) an d n e g a t i v e in A l N (ab o u t - 0.17 t o - 0.3 e V in d i f f er e n t c a l c u l a t i o n s), so i t c han g es s i g n a t an in t er m e d ia t e a l u mini u m f r a c t i o n$ $Δ_{so} : s p in - or bi t s pl i tt in g, s ma l l in t h e ni t r i d es$ $D O P : d e g r ee o f p o l a r i z a t i o n o f t h e e mi ss i o n$ $I_{T E}, I_{T M} : in t e n s i t i es o f t h e t r an s v er se - e l ec t r i c an d t r an s v er se - ma g n e t i c co m p o n e n t s o f t h e e mi ss i o n$ $η_{T E} : f r a c t i o n o f t h e e mi ss i o n t ha t i s T E p o l a r i z e d$ $(co m p r ess i v e s t r ain in t h e w e l l an d t h e q u an t u m co n f in e m e n t s hi f t t h e or d er in g an d m o v e t h e cr osso v er co m p os i t i o n, an d t h e T M e mi ss i o n l e a v es t h e c hi p t h r o ug h t h e s i d e w a l l s an d n o t t h r o ug h t h e t o p s u r f a ce)$
Mg acceptor ionization and hole density	$p \approx \frac{N _{V} N _{A}}{g _{A}} exp (- \frac{E _{A}}{2 k _{B} T})$	$p \approx \frac{N _{V}}{g _{A}} \frac{N _{A} - N _{D}}{N _{D}} exp (- \frac{E _{A}}{k _{B} T})$ $N_{V} = 2 (\frac{2 π m _{h}^{*} k _{B} T}{h ^{2}})^{3/2}$	$p : f r ee h o l e d e n s i t y a t t e m p er a t u r e T$ $N_{V} : e f f ec t i v e d e n s i t y o f s t a t es o f t h e v a l e n ce ban d$ $N_{A}, N_{D} : ma g n es i u m a cce pt or d e n s i t y an d co m p e n s a t in g d o n or d e n s i t y$ $g_{A} : d e g e n er a cy f a c t or o f t h e a cce pt or l e v e l, t ak e n a s f o u r$ $E_{A} : i o ni z a t i o n e n er g y o f t h e a cce pt or, ab o u t 160 m e V f or M g in G a N, ab o u t 320 m e V in A l_{0.7} G a_{0.3} N an d ab o u t 0.5 e V in A l N (t h e l e v e l g e t s d ee p er a s t h e a l u mini u m f r a c t i o n r i ses)$ $k_{B}, T : B o l t z mann co n s t an t an d t e m p er a t u r e$ $m_{h}^{*} : d e n s i t y - o f - s t a t es e f f ec t i v e ma ss o f t h e h o l es$ $h : P l an c k co n s t an t$ $(t h e f i r s t e x p r ess i o n i s f or t h e p a r t ia l i o ni z a t i o n o f an u n co m p e n s a t e d a cce pt or, t h e seco n d f or a co m p e n s a t e d o n e w i t h N_{D} m u c h l a r g er t han p, an d b o t h a ppl y in t h e r e g im e w h er e p i s f a r b e l o w N_{A})$
Absorption in the layers above the active region and polarization-dependent extraction	$T = exp (- α d)$	$η_{e x t r} \approx η_{T E} η_{e x t r, T E} + (1 - η_{T E}) η_{e x t r, T M}$ $η_{e x t r, T M} ≲ 0.1 η_{e x t r, T E}$	$T : f r a c t i o n o f t h e l i g h t t ha t p a sses t h r o ug h an ab sor bin g l a y er a t n or ma l in c i d e n ce$ $α : ab sor pt i o n coe f f i c i e n t o f t h e l a y er a t t h e e mi ss i o n w a v e l e n g t h (t h e p - t y p e G a N co n t a c t l a y er ab sor b s s t r o n g l y b ec a u se i t s g a p i s b e l o w t h e p h o t o n e n er g y in t h e u l t r a v i o l e t - C)$ $d : t hi c k n ess o f t h e l a y er$ $η_{e x t r} : e x t r a c t i o n e f f i c i e n cy o f t h e c hi p f or t h e mi x e d p o l a r i z a t i o n$ $η_{T E} : f r a c t i o n o f T E e mi ss i o n f r o m t h e p o l a r i z a t i o n r o w$ $η_{e x t r, T E}, η_{e x t r, T M} : e x t r a c t i o n e f f i c i e n c i es o f t h e T E an d t h e T M co m p o n e n t s t h r o ug h t h e c h ose n s u r f a ce (t h e T M l i g h t i s e x t r a c t e d m u c h l ess e f f i c i e n tl y t h r o ug h t h e t o p or t h e s u b s t r a t e, b ec a u se i t t r a v e l s in t h e pl an e o f t h e l a y er s an d i s g u i d e d)$ $(t h e g e n er a l e x t r a c t i o n e x p r ess i o n s w i t h t h e esc a p e co n e a r e o n t h e L i g h t E mi tt in g D i o d e p a g e)$

Important Questions

Ques: Why is the efficiency of the ultraviolet-C LED so much lower than that of the blue LED?

The external quantum efficiency is the product of the injection, the radiative and the extraction efficiencies (the LED page), and all three are worse in the deep ultraviolet. The injection suffers from the poor p-type doping and the hole injection, the radiative efficiency from the high threading dislocation density of AlGaN grown on sapphire and from the weak carrier localization in the wells (a statement recalled, not checked), and the extraction from the absorption of the p-GaN contact layer, from the TM polarization of the high-Al wells and from the total internal reflection at the high-index interfaces. A recent review of deep-ultraviolet LEDs reports a single high-power device with an EQE of 9.2% and a wall-plug efficiency of 7.1% at 250 mA, quotes internal quantum efficiencies of 55.65% at 237 nm, 62% at 254 nm and 93% at 284 nm, and states that the TM light extraction is no more than one tenth of the TE one. As an illustration, a 275 nm LED (4.51 eV) with an EQE of 5% at 6 V has a wall-plug efficiency of $0.05 \times 4.51/6 = 3.8%$ . Reference: M. Kneissl, T.-Y. Seong, J. Han and H. Amano, “The emergence and prospects of deep-ultraviolet light-emitting diode technologies,” Nat. Photonics 13, 233 (2019); D.-X. Zhu, G. Wang and Y.-F. Li, “Advances in enhancing the light extraction efficiency of AlGaN-based deep ultraviolet light-emitting diodes via optical polarization,” Front. Phys. 13, 1718751 (2025); Y. Taniyasu, M. Kasu and T. Makimoto, “An aluminium nitride light-emitting diode with a wavelength of 210 nanometres,” Nature 441, 325 (2006).

Ques: Why is p-type doping the central difficulty of the AlGaN LED?

The magnesium acceptor gets deeper as the aluminium fraction rises: about 160 meV in GaN, about 320 meV in Al $_{0.7}$ Ga $_{0.3}$ N (from a Hall measurement, and the true level may be deeper), and about 0.5 eV in AlN (a level depth from photoluminescence). The hole density of the acceptor expression of the doping row depends on $exp (- E_{A} /2 k_{B} T)$ , so at 300 K the change from 160 to 320 meV lowers $p$ by a factor of 22 for the same $N_{A}$ and $N_{V}$ , and the change from 160 meV to 0.5 eV lowers it by a factor of about 700. This is why the LED uses a p-GaN contact layer, a polarization-doped or superlattice layer and a tunnel junction, and these layers absorb or complicate the extraction (the absorption row). Reference: M.L. Nakarmi, K.H. Kim, K. Zhu, J.Y. Lin and H.X. Jiang, “Mg doped Al-rich AlGaN alloys for deep UV emitters,” Proc. SPIE (2004) (Texas Tech nanophotonics group; volume and paper number not retrieved).

Ques: Why does the AlGaN LED emit TM-polarized light at a high aluminium fraction?

At the zone centre the topmost valence band in GaN has the x-y character, so the recombination emits light polarized perpendicular to the c-axis (TE), and that light goes out through the top surface. The crystal-field splitting $Δ_{c f}$ is small and positive in GaN and negative in AlN (about -0.17 eV in one set of parameters), so the band with the z character rises to the top at a high aluminium fraction and the emission becomes TM. With the quasi-cubic expressions and illustrative parameters $Δ_{c f} = + 0.01$ eV, $Δ_{so} = 0.017$ eV for GaN and $Δ_{c f} = - 0.169$ eV, $Δ_{so} = 0.019$ eV for AlN (the spin-orbit value for AlN is assumed), the topmost level is the x-y one in GaN and the z-like one $E_{2}$ in AlN. A recent DFT study places the sign change of $Δ_{c f}$ at an aluminium fraction between 0.18 (linear interpolation) and 0.41 (with bowing), and the strain in the well shifts the crossing again, so the TE-to-TM composition for a device depends on the strain, the well width and the barrier. Reference: Reference for the composition range: the DFT study “Composition dependent k.p band parameters for wurtzite (Al,Ga)N alloys from density functional theory,” arXiv:2505.19730 (2025); the quasi-cubic expressions are in S.L. Chuang and C.S. Chang, Phys. Rev. B 54, 2491 (1996) (recalled, not checked).

Ques: How uncertain is the wavelength of an AlGaN LED of a given composition?

The bowing parameter reported for the AlGaN bandgap is not settled. A fit of absorption data gave $E_{g} = 3.43 + 1.44 x + 1.33 x^{2}$ eV, which is a bowing parameter of 1.33 eV with $E_{g} (AlN) = 6.2$ eV, while other compilations use a value near 0.7 eV. For $x = 0.5$ the first value gives 4.48 eV (277 nm) and the second 4.64 eV (267 nm), a difference of 10 nm for the same composition, and for $x = 0.3$ and 0.7 the values are 311 and 301 nm, and 244 and 237 nm. The strain and the quantum confinement shift the emission in addition, so the composition of a device is not found from the wavelength alone. Reference: W. Shan, J.W. Ager III, K.M. Yu, W. Walukiewicz, E.E. Haller, M.C. Martin, W.R. McKinney and W. Yang, “Dependence of the fundamental band gap of Al $_{x}$ Ga $_{1 - x}$ N on alloy composition and pressure,” J. Appl. Phys. 85, 8505 (1999), DOI 10.1063/1.370696, the fit and the 6.2 eV value being those printed in that paper; for the alternative bowing value see I. Vurgaftman and J.R. Meyer, “Band parameters for nitrogen-containing semiconductors,” J. Appl. Phys. 94, 3675 (2003) (recalled, not checked).

Ques: What is the difference between the AlGaN bowing parameters of 1.33 eV and 0.7 eV, and which should be used?

The two numbers describe the same formula $E_{g} (Al_{x} Ga_{1 - x} N) = (1 - x) E_{g} (GaN) + x E_{g} (AlN) - b x (1 - x)$ with a different $b$ , so they differ only in how much the gap of the alloy lies below the straight line between GaN and AlN. With $E_{g} (GaN) = 3.43$ eV and $E_{g} (AlN) = 6.2$ eV the gap at $x = 0.5$ is 4.64 eV (267 nm) for $b = 0.7$ eV, 4.58 eV (271 nm) for $b = 0.94$ eV and 4.48 eV (277 nm) for $b = 1.33$ eV (computed). The largest difference between 0.7 and 1.33 eV is $0.63 x (1 - x) = 0.16$ eV at $x = 0.5$ , which is 9 to 10 nm in the deep ultraviolet, and it is zero at the two ends. The value 1.33 eV is the fit of a single absorption study (Shan et al.), and 0.7 eV is the value of a compilation (Vurgaftman and Meyer, recalled). A review of the more recent literature gives a range of 0.86 to 1.00 eV for the constant bowing parameter and recommends 0.94 eV, and the spread is attributed to the sample quality, the accuracy of the composition and the strain state of the samples. Another compilation (Wu, 2009) lists 0.6 eV. So the reported values run from 0.6 to 1.33 eV, and the middle of the range (0.7 to 1.0 eV) is the most commonly used. The value 1.33 eV is at the high end and is tied to the AlN gap of 6.2 eV used in the same fit. For the design of a device the safe procedure is to compute the wavelength with both ends of the range, to treat the spread of 10 nm as an uncertainty, and to calibrate the composition by X-ray diffraction and not from the wavelength. Reference: W. Shan et al., J. Appl. Phys. 85, 8505 (1999) (the 1.33 eV fit, see the previous question); the 0.86 to 1.00 eV range and the 0.94 eV value are from the open-access article at pmc.ncbi.nlm.nih.gov/articles/PMC7530746 (the recommendation is attributed there to Coughlan et al., 2015, which I did not read); J. Wu, J. Appl. Phys. 106, 011101 (2009), Table III (0.6 eV); I. Vurgaftman and J.R. Meyer, J. Appl. Phys. 94, 3675 (2003) for the 0.7 eV value (recalled, the text could not be read).

Quartz 5

Explorer