Photoluminescence (PL) spectra from ruby were obtained using a highly stable LED light source, employing pulse width modulation technique for excitation. The temporal variation in PL intensity caused by the increasing temperature of the LED used for excitation can be mitigated by adjusting the duty ratio (%) of the pulsed LED light to below 10% for cooling the LED. Stable PL spectra measurements were achieved with a duty ratio of less than 10% using a duty ratio-controlled pulsed LED light source, as temperature fluctuations in LED light intensity are minimized at duty ratios less than 10%. Furthermore, fluctuations in the measured PL intensity were diminished by setting the frequency of the pulsed LED light source to greater than 1 kHz. This method enables more reliable, cost-effective, and stable PL measurements for material characterization in semiconductors, photonics, and nanotechnology.

Light-emitting diodes (LEDs) have emerged as a crucial light source in photoluminescence (PL) measurements, offering benefits such as brightness, a wide range of colors (spanning peak wavelengths from UVC to IR region), high efficiency, low power consumption, and compactness.1,2 Despite these advantages, commercially available LEDs face drawbacks such as a reduction in luminescence intensity and a shift to longer emission wavelengths upon operation,3–10 primarily because of the temperature increases caused by illumination.

Various techniques have been developed to improve the increase in peak wavelength of the LED light source and the peak intensity decreasing as the temperature increases with LED illumination. Temperature control technique has been proposed3–10 using a cooling device to suppress the rise in temperature caused by the powered state of the LED. Feedback control to adjust the input current of the LED for compensating the luminescence intensity fluctuations and preconditioning the LED light source to its final (equilibrium) temperature using a furnace are considered effective techniques to stabilize the LED light source. However, these methods often require prolonged periods to reach equilibrium and require additional apparatuses, such as furnaces and temperature controllers, compromising the LED benefits, such as simplicity, compactness, and energy efficiency.

To stabilize the optical intensity of conventional LED light sources, conventional practices involved aging LEDs for ∼30 min before utilization or operating them at lower input currents than their rated capacity to minimize the increase in temperature caused by energization. This study reports on the advancements in LED light source development focusing on PL measurement applications. A highly stable LED light source has been devised, which fulfills the requirements of a duty ratio-controlled pulsed LED based on pulse width modulation (PWM) technique at a small duty ratio for effective cooling of the LED.

Ruby crystals with a concentration of 0.4 mol. % Cr2O3 were synthesized using the floating zone (FZ) technique using highly pure Al2O3 (99.99%) and Cr2O3 (99.9%) powders as the starting materials. The quality of the grown crystals was assessed via x-ray powder diffraction and photoluminescence (PL) spectra. X-ray powder diffraction patterns were obtained using a RINT ULTIMA instrument (Rigaku Co., Japan).

PL spectra were acquired employing a duty ratio-controlled pulse LED light source, based on pulse width modulation (PWM) technique, and an optical fiber monochromator (USB-2000, Ocean Optics Co., USA). In addition, a Si photodiode (PDA36A, Thorlabs, USA) was utilized to measure the PL intensity from ruby and the luminescence intensity of the LED light source. The signal intensities from the Si photodiode were recorded as true RMS values using a digital multimeter (PC7000M, Sanwa Co., Japan) and a lock-in amplifier (LI5460, NF Co., Japan). The peak intensity values, Vp–p, were also measured using USB oscilloscopes—PicoScope 2204 and PicoScope 4424 (Pico Technology Co., UK).

Figure 1 presents a schematic illustration of the photoluminescence evaluation equipment excited using a duty ratio-controlled pulse LED light source (λ = 470 nm) employing the PWM technique. The PWM signals were generated using WF1945, WF1943A, DF-1905 (NF Co., Japan), and AFG-2125 (Instek Co., Japan).

FIG. 1.

Schematic of photoluminescence evaluation equipment excited with a duty ratio-controlled pulse LED light source (λ = 470 nm). PL intensity is measured using true rms (root mean square) and/or peak intensity of pulse (Vp–p).

FIG. 1.

Schematic of photoluminescence evaluation equipment excited with a duty ratio-controlled pulse LED light source (λ = 470 nm). PL intensity is measured using true rms (root mean square) and/or peak intensity of pulse (Vp–p).

Close modal

For a portable pulsed light source, the PWM generator board XY-LPWM was utilized. The LED drivers included NJW4617DL3, NJW4616U2, or NJU6080F1 (Nisshinbo Micro Devices Co., Japan), and a DC constant current source (Takasago Co., Japan) served as the constant current pulse LED driver for DC constant illumination of the LED. The duty ratio of the LED light source was carefully adjusted to stabilize the temporal fluctuations in optical intensity during illumination.

PL spectra and intensities were measured based on true rms (root mean square) values and/or peak intensity of the pulse (Vp–p) according to the following equations:
Duty%=100τσ+τ,
(1)
true RMS=1T0Txt2dt,
(2)
true RMS=Vppτσ+τ=VppDuty,
(3)
where Duty% denotes the percentage of duty ratio, τ indicates the lighting period of LED, σ represents the length of time LED is on, true rms is the root mean square of pulse signal, T denotes the length of the cycle, x(t) represents the signal intensity of pulse at time t, and Vp–p indicates the peak-to-peak intensity of the pulse signal. In the field of pulse signal intensity measurement, true rms is extensively utilized across various measurement systems because of its convenience. The intensity of a pulse signal measured in the true rms varies with the duty cycle percentage (Duty%) according to Eq. (3), unlike the peak-to-peak voltage (Vp–p), which remains constant regardless of the fluctuations in duty cycle.

Figure 2 illustrates the temporal variations in input currents, LED temperatures, and optical intensities from LEDs, along with the optical intensities from a duty ratio-controlled pulse LED light source at duty ratios of 90%, 50%, and 10%. Both pulse peak intensities (Vp–p) and true rms values are depicted in the figure. In continuous LED illumination scenarios in the lower portion of Fig. 2(a), the optical power from the LED diminishes, and the peak wavelength of the LED light gradually increases beyond 10–30 min due to the temperature increases caused by illumination. Conversely, in the pulse lighting mode exhibited in the upper portion of Fig. 2(a), the LED is cooled during off-periods, thereby reducing temporal fluctuations in optical power and peak wavelength.

FIG. 2.

Schematic of temporal variations of input current, LED temperature, and optical intensity from LED in the continuous lighting mode (lower portion) and the pulse lighting mode (upper portion) (a). Optical intensities from a duty ratio-controlled pulse LED light source at Duty% = 90% (b), 50% (c), and 10% (d). Pulse peak intensities Vp–p and true rms (root mean square) are also plotted.

FIG. 2.

Schematic of temporal variations of input current, LED temperature, and optical intensity from LED in the continuous lighting mode (lower portion) and the pulse lighting mode (upper portion) (a). Optical intensities from a duty ratio-controlled pulse LED light source at Duty% = 90% (b), 50% (c), and 10% (d). Pulse peak intensities Vp–p and true rms (root mean square) are also plotted.

Close modal

LED cooling is primarily governed by thermal conduction, dependent on the temperature differential between the LED and the ambient temperature. The temperature of the LED is influenced by factors such as the input current, illumination duration, heat generation efficiency, and effective specific heat capacity of the LED. The temperature increases until the heating and cooling processes reach equilibrium, typically within 10–30 min in the continuous lighting mode shown in the lower portion of Fig. 2(a). On the other hand, in the pulse lighting mode in the upper portion of Fig. 2(a), the time for the temperature equilibrium varies depending on the duty ratio.

In the high-frequency pulse lighting scenarios depicted in Figs. 2(b)2(d), both pulse intensity (Vp−p) and true rms values gradually decreased at higher duty cycles—90% [Fig. 2(b)] and 50% [Fig. 2(c)]. In the pulse lighting LED in Figs. 2(b)2(d), the value of t-rms is smaller than that of peak intensity Vp−p by the amount of Duty% according to Eq. (3). Conversely, at a lower duty cycle of 10% [Fig. 2(d)], the pulse intensity (Vp−p) and true rms values remain stable due to effective cooling during the longer off-periods, which stabilizes the LED temperature. This study experimentally confirmed that temporal fluctuations in optical intensity from LED light sources can be successfully stabilized using the PWM technique at Duty% < 10%. In the pulse lighting scenario with a duty ratio of 10% or less shown in Fig. 2(d), temporal changes in the t-rms value and Vp−p value are extremely stable and changes are small. Stable fluorescence measurement is, therefore, possible by using PWM modulation and pulse lighting of the LED with a duty ratio of 10% or less. Pulse LEDs driven at a small duty ratio can be used as a stable light source even with a large input current. This light source is considered advantageous for measuring weak fluorescence.

Figure 3 displays the PL spectra from 0.4 mol. % Cr doped ruby excited with a duty ratio-controlled pulse LED light source (λ = 470 nm), revealing a PL peak from ruby at λ = 694 nm (R line). Typical photoluminescence (PL) equipment measures using the average value or the integral value of PL over a certain period. The t-rms value may also be used. As shown in the inset of Fig. 3, these measured values are in a good proportional relationship with duty ratio. This demonstrates that PL spectra can be accurately measured using a pulse LED light source for excitation, with high stability achieved at duty ratios below 10%. The PL peak intensity at λ = 694 nm varies from 1.0% to 50%, as displayed in Fig. 3, indicating that the PL intensity measured in true rms increases with Duty% as per Eq. (3).

FIG. 3.

PL spectra from 0.4 mol. % Cr doped ruby excited with a duty ratio-controlled pulse LED light source (λ = 470 nm). Variation of PL peak intensity at λ = 694 nm (R line) with Duty% of pulse LED light source (λ = 470 nm) is shown in the figure. PL intensity is measured using true rms (root mean square).

FIG. 3.

PL spectra from 0.4 mol. % Cr doped ruby excited with a duty ratio-controlled pulse LED light source (λ = 470 nm). Variation of PL peak intensity at λ = 694 nm (R line) with Duty% of pulse LED light source (λ = 470 nm) is shown in the figure. PL intensity is measured using true rms (root mean square).

Close modal

Figure 4 illustrates the temporal fluctuations of PL peak intensities at λ = 694 nm (R line) from 0.4 mol. % Cr doped ruby, excited using a pulse LED light source (λ = 470 nm) at various values of Duty% (1.0%–50%). The magnified views at Duty% = 1.0% and 50% are presented in the figure as well, displaying that PL intensities from ruby gradually decrease after lighting at Duty% > 30% but become stable at Duty% < 10%.

FIG. 4.

Temporal variations of PL peak intensities at λ = 694 nm (R line) from 0.4 mol. % Cr doped ruby excited using a pulse LED light source (λ = 470 nm) at various duty ratios from 1.0% to 50%. Magnified plots at Duty% = 1.0% and 50% are depicted as well.

FIG. 4.

Temporal variations of PL peak intensities at λ = 694 nm (R line) from 0.4 mol. % Cr doped ruby excited using a pulse LED light source (λ = 470 nm) at various duty ratios from 1.0% to 50%. Magnified plots at Duty% = 1.0% and 50% are depicted as well.

Close modal

Figure 5 depicts the PL peak intensity at λ = 694 nm (R line) from 0.4 mol. % Cr doped ruby, excited using a pulse LED light source (λ = 470 nm) at Duty% = 50%. The stability of PL peak intensities from ruby was affected by the pulse frequencies of the LED light source within the range of 10 Hz to 10 kHz, with fluctuations observed at frequencies lower than 1 kHz. Stable measurements of PL spectra and intensities are achievable using a pulse LED light source based on a PWM technique at small Duty% of 10% or less.

FIG. 5.

PL peak intensity at λ = 694 nm (R line) from 0.4 mol. % Cr doped ruby excited using a pulse LED light source (λ = 470 nm) at Duty% = 50%, at various frequencies.

FIG. 5.

PL peak intensity at λ = 694 nm (R line) from 0.4 mol. % Cr doped ruby excited using a pulse LED light source (λ = 470 nm) at Duty% = 50%, at various frequencies.

Close modal

PL spectra and intensities can be effectively measured using a pulse LED light source utilizing a PWM technique at low Duty%. The fluctuations in optical intensities of the LED can be significantly reduced with pulse lighting at duty ratios below 10%, where the extended cooling time afforded by low Duty% enhances the stabilization of the LED light source. Consequently, a duty ratio-controlled pulse LED light source is anticipated to serve as an exceptionally stable light source for optical measurements, including photoluminescence, optical scattering, reflection, and absorption.

The authors thank Yoshimitsu Muto, Yuga Kinoshita, Daiki Takada, Yohei Murakami, Koki Nasu, and Ayaka Kojima of Crystal Engineering Laboratory, Toyo University, for their assistance in the experiments. The authors are also grateful to Takeshi Matsumoto, Matsumoto Precision Co., for his help in the pulsed LED light system development. In addition, the authors acknowledge Enago (www.enago.jp) for English language editing and proofreading.

The authors have no conflicts to disclose.

A.M. and T.K. contributed equally to this work.

Ami Hitomi: Conceptualization (equal); Data curation (equal); Writing – review & editing (equal). Toru Katsumata: Conceptualization (equal); Data curation (equal); Writing – original draft (equal); Writing – review & editing (equal). Hiroaki Aizawa: Supervision (equal).

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

1.
T.
Katsumata
,
N.
Hanami
, and
H.
Aizawa
,
Rev. Sci. Instrum.
92
(
11
),
114903-1
-
114903-8
(
2021
).
2.
H.
Aizawa
,
Y.
Miyazaki
,
T.
Katsumata
, and
S.
Komuro
,
J. Electrochem. Soc.
168
(
1
),
017510-1
-
017510-6
(
2021
).
3.
D.
Lee
,
H.
Choi
,
S.
Jeong
,
C. H.
Jeon
,
D.
Lee
,
J.
Lim
,
C.
Byon
, and
J.
Choi
,
Int. J. Heat Mass Transfer
127
,
1243
(
2018
).
4.
P.
Fredes
,
U.
Raff
,
E.
Gramsch
,
J.
Pascal
, and
J.
Cuenca
,
Microelectron. Reliab.
98
,
24
30
(
2019
).
5.
M. J.
Kalani
,
M. S.
Salay Naderi
, and
G.
B Gharehpetian
,
Comput. Electr. Eng.
77
,
191
204
(
2019
).
6.
R.
Singh
,
M.
Mochizuki
,
T.
Yamada
, and
T.
Nguyen
,
Appl. Therm. Eng.
166
,
114733-1
-
114733-7
(
2020
).
7.
K.
Delendik
,
N.
Kolyago
, and
O.
Voitik
,
Comput. Math. Appl.
83
,
84
94
(
2021
).
8.
J.
Hegedüs
,
G.
Hantos
, and
A.
Poppe
,
Microelectron. Reliab.
79
,
448
456
(
2017
).
9.
X.
Lin
,
S.
Mo
,
L.
Jia
,
Z.
Yang
,
Y.
Chen
, and
Z.
Cheng
,
Appl. Energy
242
,
232
238
(
2019
).
10.
X.
Lin
,
S.
Mo
,
B.
Mo
,
L.
Jia
,
Y.
Chen
, and
Z.
Cheng
,
Appl. Therm. Eng.
172
,
115165-1
-
115165-8
(
2020
).