TUGAS I

METODE ANALISIS BAHAN

Oleh :
Anis Silfiyatus Sa’adah
01111940000016

DEPARTEMEN FISIKA

FAKULTAS SAINS DAN ANALITIKA DATA

INSTITUT TEKNOLOGI SEPULUH NOPEMBER SURABAYA

TAHUN 2022

Pengamatan in-situ grafena menggunakan Mikroskop Optik
Mikihiro Kato, Sujun Guan, Xinwei Zhao (2021)
Tujuan :
1. Mengamati kelayakan pengamatan in-situ
dekomposisi grafena dengan menggunakan OM
(Optical Microscope) yang dilengkapi dengan tahap
vakum pemanas
2. Mengamati perilaku multi-layers graphene (MLG)
pada substrat logam
3. Mengamati pengaruh film graphene yang
ditumbuhkan CVD (Chemical Vapor Deposition)
pada foil logam selama pemanasan dalam vakum dan
atmosfer gas Ar + H2.
Hasil :
1. Pengamatan graphene dari Ni foil
Gambar1(a) menunjukkan gambar OM dari sampel
MLG/Ni di RT. Sampel ini dipanaskan dari RT sampai
9000C, dengan laju 200C/menit. Gambar1(c), di atas
7500C, mulai terjadi perubahan di wilayah I dan II,
dilingkari hitam. Reaksi pengendapan dan segregasi terjadi
pada 7250C hingga 9000C. Perubahan ini disebabkan oleh
karbon terlarut dalam Ni mulai terdeposisi kembali atau
MLG permukaan mulai larut menjadi Ni, selama proses
pemanasan. Gambar 1 (5) - (6), perubahan permukaan di
wilayah III lebih jelas daripada di gambar suhu yang lebih
rendah, dan perbedaan perubahan di tiga wilayah
kemungkinan karena orientasi kristal dan permukaan
bentuk Ni.
Gambar 2. Menunjukkan hasil OM in-situ dari grafena
yang ditumbuhkan CVD pada permukaan sampel Ni
selama kenaikan suhu di atmosfer Ar + H2. Ketika sampel
dipanaskan dalam Ar + H2, area berwarna biru tua dan
oranye muncul pada permukaan sampel pada Gambar 2.
Proses pereaksian ini berbeda dengan proses pereaksian
pada kondisi vakum. Daerah ini adalah karbon amorf (aC).
Setengah dari area grafena tergores oleh gas Ar + H2.
1
Dengan mempertahankan sampel pada 9000C dalam atm,
grafena hene tergores secara bertahap.
2. Pengamatan graphene dari Cu foil
Gambar 3 menunjukkan bahwa perubahan film grafena
yang ditumbuhkan CVD pada foil Cu selama dipanaskan
dalam vakum dan atmosfer gas Ar + H2 oleh OM in-situ
pengamatan. Gambar 3a hampir tidak menunjukkan
perubahan warna sampai suhu hingga 4000C selama
dipanaskan dalam ruang hampa. Dengan meningkatnya
suhu hingga 5000C, perubahan warna terlihat jelas.
Sedangkan dengan pemanasan lebih lanjut di atas 8000C,
perubahannya sedikit. Perubahan warna pada citra OM
disebabkan oleh dekomposisi grafena, yang terkait dengan
pelepasan gas interkalasi, seperti oksigen residu antara
logam dan grafena, dan/atau penghilangan lapisan oksida.
Sedangkan perubahan film grafena yang ditumbuhkan
CVD pada Cu foil selama dipanaskan dalam atmosfer gas
Ar + H2 ditunjukkan pada Gambar 3b. Grafena stabil
dengan suhu di bawah 4000C. Dengan meningkatnya suhu
hingga 7000C, warna grafena menjadi bagian dari hijau,
yang disebabkan oleh pengikisan oleh atmosfer Ar + H2
pada suhu tinggi. Dengan pemanasan lebih 9000C,
perubahannya jelas, yang mengisyaratkan bahwa film
graphene telah sepenuhnya tergores.
Kesimpulan :
Perubahan MLG pada Ni dan grafena pada Cu dapat
dikenali melalui citra mikroskop optik. Pengamatan in-situ
dekomposisi grafena telah dilakukan dengan menggunakan
mikroskop optik yang dilengkapi dengan ruang vakum
pemanas. Karakteristik perubahan permukaan tersebut
sesuai dengan kondisi pertumbuhan grafena sebagai
katalisator Ni dan Cu. Puncak G-band bergeser ke sisi
energi yang lebih rendah, dengan meningkatnya suhu
panas dalam ruang hampa dan atmosfer gas Ar + H2.
Metode ini sederhana dan cukup efektif untuk memperoleh
informasi tentang kondisi pertumbuhan area grafena yang
luas pada substrat seperti Ni, yang memiliki tingkat
kelarutan karbon yang tinggi. Mikroskop optik dapat
mengamati grafena pada logam dengan mudah. Hasil
pengamatan menunjukkan secara nyata dengan gambar
berwarna dan akan memberikan cara sederhana untuk
memahami keadaan grafena dalam berbagai kondisi
atmosfer.
 2
Dehidroksilasi dan kristalisasi gelas: Studi DTA
D.C. Lago, M.O. Prado (2013)
Differential Thermal Analysis (DTA) adalah teknik yang
sering digunakan untuk mempelajari kinetika kristalisasi
serbuk kaca selama sintering. Penelitian ini bertujuan
untuk menganalisis bagaimana puncak endotermik
dehidroksilasi yang muncul pada pengukuran DTA
bergantung pada area spesifik dan pada laju pemanasan.
Penelitian ini dilakukan pada gelas yttrium aluminosilicate
(YAS), samarium aluminosilicate dan soda-lime silica,
slab dan powder.

Gambar 1 menunjukkan hasil DTA untuk monolit dan

fraksi yang berbeda dari bubuk kaca YAS dalam cawan
lebur alumina. Jejak yang sesuai dengan pelat
menunjukkan infleksi Tg dan puncak kristalisasi seperti
yang diharapkan (menjadi TS suhu yang sesuai dengan
awal puncak kristalisasi dan TX, suhu puncak
maksimum).
Sedangkan Gambar 5 menunjukkan jejak DTA untuk
Hasil analisis DTA disajikan kembali pada Tabel 1 lempengan dan bubuk NCS dimana tidak ada puncak
endotermik yang diamati.

Jadi, reaksi endotermik permukaan ini, dengan entalpi

reaksi nyata sebesar 255 kkal mol-1 untuk kaca YAS, dapat
digolongkan ke dalam reaksi dehidroksilasi.

Gambar 2 menunjukkan tumpang tindih jejak DTA Kesimpulan :

untuk fraksi YAS 3 dan luas penampang silinder bubuk
selama sintering dalam hot stage microscope (HSM).
Penyusutan terjadi sebelum puncak endotermik.
Karena puncak eksotermik DTA diamati dengan bubuk
kaca YAS, tetapi tidak dengan pelat, asal fisik puncak ini
harus dikaitkan dengan sifat permukaan kaca. Untuk
mengeksplorasi fakta ini, pada Gambar 3 dibandingkan
jejak DTA untuk fraksi bubuk kaca YAS 2, fraksi bubuk
kaca YAS 2 + 30 jam dalam air suling pada suhu kamar
dan bubuk kaca YAS yang dibulatkan dalam nyala
propana-oksigen.
Puncak endotermik didefinisikan dengan baik untuk
bubuk, tetapi hanya disarankan untuk bubuk yang kontak
dengan air atau yang dilewatkan melalui nyala propana-
oksigen.
Gambar 4 menunjukkan hasil DTA pelat kaca
aluminosilikat samarium.
Dalam mempelajari kristalisasi serbuk kaca, terdapat
pengaruh pergerakan puncak kristalisasi menuju suhu yang
lebih rendah, jika dibandingkan dengan hasil yang
diperoleh dengan lempengan kaca. Terlebih lagi ketika
dijalankan pada beberapa serbuk kaca, gelas YAS
misalnya, jejak DTA menunjukkan puncak endotermik
sebelum puncak kristalisasi. Puncak endotermik ini
muncul pada serbuk kaca. Puncak endotermik ini emiliki
tinggi, lebar dan rentang suhu yang tergantung pada sifat
kimia permukaan kaca. Karena prosesnya endotermik, ini
bukan karena relaksasi regangan, kristalisasi atau energi
yang tersedia karena permukaan penurunan sintering.
Sintering dapat menunjukkan perubahan pada baseline
DTA tetapi bukan puncak endotermik.

Pengukuran XRF yang diinduksi laser Femtosecond dari senyawa protein-logam
Jungu Kang, Do-Kyeong Ko (2001)
Penelitian ini mengukur XRF senyawa logam-protein
oleh sumber sinar-X yang diinduksi laser dengan target
pita aluminium. Sinar-X yang diinduksi laser dihasilkan
oleh tingkat pengulangan tinggi (1 kHz) femtosecond (40
fs) Untuk mengkarakterisasi sumber sinar-X K𝛼 dan
bremsstrahlung yang diinduksi laser, penelitian ini
mengubah intensitas laser dan posisi fokus pulsa. Selain
itu, dilakukan juga mengukur sinyal XRFdengan pelat
tembaga, bubuk selenium, hemoglobin terliofilisasi, dan
seleno-metionin terliofilisasi, yang dideteksi oleh
spektrometer sinar-X. Dengan beberapa mikrogram
sampel, XRF dapat diukur dengan akurasi tinggi dan
stabilitas tinggi.
Gambar 1 menunjukkan pengaturan eksperimental
untuk pembangkitan sinar-X yang diinduksi laser dan
pengukuran fluoresensi yang memiliki rentang energi yang
luas, dari 0,1 keV hingga 20 keV.
Pulsa laser femtosecond difokuskan pada intensitas
1016 W/cm2 atau lebih oleh cermin parabola off-axis.
Sampel senyawa logam ditempatkan di belakang sumber
sinar-X yang diinduksi laser, dan fluoresensi logam diukur
dengan menggunakan detektor sinarX Amptek (XR-
100CR). Untuk mengurangi kebisingan, pelat Pb dililitkan
di sekitar detektor sinar-X untuk memblokir sumber sinar-
X yang diinduksi laser dan mengoptimalkan sinyal
fluoresensi dengan menyesuaikan posisi dan kecepatan
target pita Al. XRF yang diinduksi laser dari pelat tembaga
(Cu), moglobin manusia (Fe), bubuk selenium (Se), dan
seleno-peptida (Se) diukur untuk meneliti fluoresensi
sinar-X dalam berbagai rentang energi.
3
Gambar 1. (a) Spektrum sinar-X yang diinduksi laser
dengan aluminium. Ini sesuai dengan intensitas laser 1016
W/cm2 dan sudut datang 30𝛼. (b) Ketergantungan fluks
aluminium K𝛼 pada daya laser femtosecond. (c)
Ketergantungan nomor foton aluminium K𝛼 (1,486 keV)
dan emisi bremsstrahlung (>3,5 keV) pada daya laser
femtosecond. (d) Efisiensi konversi perubahan K𝛼
aluminium yang diinduksi laser oleh daya laser.
Gambar 2. Spektrum XRF yang diinduksi laser dari
bubuk selenium dan pelat tembaga (a). Stabilitas XRF yang
diinduksi laser dengan bubuk selenium (b)
Gambar 3. Menunjukkan Spektrum XRF yang diinduksi
laser dari bubuk hemoglobin manusia yang diliofilisasi (a)
dan spektrum XRF yang diinduksi laser dari bubuk seleno-
metionin yang diliofilisasi (b)
Kesimpulan :
Dalam penelitian ini, XRF dari berbagai senyawa logam
adalah diukur dengan sumber sinar-X yang diinduksi laser.
Dengan menggunakan wilayah energi bremstrahlung lebar
dari sumber sinar-X yang diinduksi laser, dimungkinkan
untuk menganalisis XRF dari berbagai senyawa logam
yang memiliki karakteristik energi sinar-X yang spesifik.
Untuk mengoptimalkan intensitas sumber sinar-X yang
diinduksi laser, dapat mengubah lebar pulsa, daya laser,
dan posisi pemfokusan pulsa. Tren emisi Al Kÿ dan
bremsstrahlung diukur sesuai dengan perubahan
intensitas laser untuk menemukan kondisi efisiensi
konversi yang optimal,
dan perbedaan jarak 10–20 m antara posisi maksimum
Al K𝛼 dan posisi maksimum bremsstrahlung dikonfirmasi.
Peneliti mengukur XRF dengan pelat tembaga, bubuk
selenium, hemoglobin terliofilisasi, dan seleno-metionin
terliofilisasi. Seperti kandungan atom logam dalam
komposisi logam analisis, jumlah foton per detik XRF
berkurang, dan rasio signal-to-noise serta akurasi rasio
K𝛼/K juga menurun. Telah dipastikan bahwa pengukuran
XRF juga dimungkinkan dengan beberapa miligram
senyawa logam. Sinyal energi sinar-X emisi karakteristik
akurat diukur dengan sejumlah kecil atom logam.
 Applied Surface Science Advances 6 (2021) 100138

Contents lists available at ScienceDirect

Applied Surface Science Advances

journal homepage: www.sciencedirect.com/journal/applied-surface-science-advances

In-situ observation of graphene using an optical microscope

Mikihiro Kato a, *, Sujun Guan b, Xinwei Zhao a, *
a
Department of Physics, Tokyo University of Science, 1-3 Kagurazaka, Shinjuku-ku Tokyo 162-8601, Japan
b
Bio-nano Electronics Research Center, Toyo University, Kawagoe, Saitama 350-8585, Japan

A R T I C L E I N F O A B S T R A C T

Keywords: In-situ observation of graphene by using visible light was very few. Herein, we investigate the feasibility of in-situ
In-situ observation observation of graphene decomposition by using the optical microscope (OM) equipped with a heating vacuum
Graphene stage. In this work, in addition to in-situ observation, we investigated the behavior of CVD-grown graphene on
Optical microscope
metal (Ni and Cu) surfaces using in-situ Raman spectroscopy and X-ray photoelectron spectroscopy to understand
CVD
the surface changes observed in the OM images. Observations were also made in vacuum and argon hydrogen gas
flow conditions. As a result of the observation, it was possible to observe the decomposition of graphene using
visible light at high temperatures. We expect it to be an easy way to observe graphene under various atmospheric
conditions.

1. Introduction necessary to simply equipment observe the behave of graphene under

Graphene has been in the spotlight because of its fascinating prop­
erties, which are ideal for micro-and nanomechanical systems, thin film
transistors, transparent and conductive composites, and electrodes,
flexible and printable (Opto) electronics, and photonics [1,2]. Graphene
has been widely studied for its fabrication and controllability. In
particular, the chemical vapor deposition (CVD) method is the main­
stream graphene growth process for graphene nowadays [3–7]. CVD
synthesis is considered a promising method to fabricate 2D materials
such as graphene because of its ability to mass produce a high-quality
thin film. It reported that the thermal CVD method combined with the
roll-to-roll process [8], in which a continuous graphene synthesis was
realized on a long copper (Cu) foil. The growth process of graphene is
well known as using a nickel (Ni) or Cu catalyst. For the case of using the
Ni substrate, carbon sources first dissolved in Ni at high temperatures.
After that, graphene grows on the Ni surface during the cooling process
[9–11]. On the other hand, the use of Cu substrate enables a monolayer
graphene formation, which is caused by the low carbon solubility in Cu
at high temperatures [12–14]. These mechanisms have been revealed by
in-situ observation, such as by using low energy electron microscopy
(LEEM), scanning electron microscope (SEM) and transmission electron
microscope (TEM) [15–19]. These methods can be used for in-situ
observation at the atomic level [20,21], but they require large equip­
ment. In addition, the observation environment is a high vacuum, which
makes it difficult to observe under various condition. Therefore, it is
various condition. Furthermore, it is difficult to flexibly combine ana­
lyzers such as Raman spectroscopes into the measurement system.
Therefore, optimization of growth conditions and properties is generally
done by analyzing graphene "after CVD synthesis" with Raman spec­
troscopy and other methods.
Recently, the observation of the growth of graphene by using an
optical microscope (OM) based on black body radiation was reported
[22,23]. The in-situ observation by using visible light, however, was
very few.
In this paper, we investigate the feasibility of in-situ observation of
graphene decomposition by using the OM equipped with a heating
vacuum stage. The behavior of the multi-layers graphene (MLG) on
metal substrates observed. Moreover, this work is carried out in the
investigation of the influence of the CVD-grown graphene films on metal
foil during heated in vacuum and Ar + H2 gas atmosphere, respectively,
by a portable Raman spectroscopy system combining with the in-situ
OM. In situ observation showed changes in the optical image in the
temperature range characteristic of graphene CVD synthesis. The change
of the graphene films has been directly observed by the in-situ OM and
Raman observation. The relationship between the change of the gra­
phene films and the heated temperature in vacuum and Ar + H2 gas
atmosphere is discussed.

* Corresponding authors.
E-mail addresses: 1214702@alumni.tus.ac.jp (M. Kato), xwzhao@rs.kagu.tus.ac.jp (X. Zhao).

https://doi.org/10.1016/j.apsadv.2021.100138
Received 27 March 2021; Received in revised form 2 July 2021; Accepted 5 July 2021
Available online 14 July 2021
2666-5239/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
M. Kato et al.
2. Experimental
2.1. CVD-grown graphene films
The starting sample was the polycrystalline Ni foil (>99%: The
Nilaco Corporation, Japan) surface coated with a thin carbon film (in the
form of disordered few layer graphene) grown by thermal CVD. The
carbon deposition was carried out with ethanol vapor supplied by
bubbling liquid ethanol with an Ar gas at 870 ◦ C for 30 min (total
pressure of 5 Torr).
In the same way, graphene films were formed on the surface of the
polycrystalline Cu foil (>99.99%: The Nilaco Corporation, Japan) at 950
ºC for 10 min by thermal CVD using bubbled ethanol as the carbon
source. Carbon deposition was conducted with ethanol vapor supplied
by bubbling liquid ethanol with an Ar gas (total pressure of 5 Torr).
2.2. Characterization
The change of the CVD-grown graphene films had been observed by
an in-situ OM instrument (KH-7700, Hirox). The magnification range of
the objective lens is from 140 to 1400 times. The light source is a metal
halide lamp. The heating stage for the in-situ observation is set in a
vacuum chamber (S.T. Japan Co., Ltd) with an observation window. The
observation window is made of quartz. The temperature was from RT to
900 ºC. In the case of vacuum heating, the pressure of the vacuum
chamber was kept better than 10− 5 Torr. While heated in Ar + H2 gas
atmosphere, the Ar + H2 gas has flowed at a rate of 30 ml/min. A
portable Raman spectroscopy system (i-Raman EX, B&W TEC) was used
for in-situ observation of Raman. The 532-nm of an Nd-YAG laser was
used for exciting the Raman scattering. The spectra of C 1 s was analyzed
by X-ray photoelectron spectroscopy (XPS, JPS-9010MC, JEOL).
3. Results and discussion
3.1. Observation of graphene of Ni foil
Fig. 1(a) shows the OM images of the MLG/Ni sample at RT. The
sample surface covered with MLG shows. This sample was heated from
RT to 900 ºC, with a rate of 20 ºC /min. It focused on the three regions I,
II and III as shown in Fig. 1(a). From Fig. 1(b), neither the surface
Fig. 1. (a) OM image of RT. From (b) to (f) In-situ OM images of graphene temperature rising.
2
Applied Surface Science Advances 6 (2021) 100138
condition nor the vacuum (3.0 × 10− 5 Torr) changed noticeably, until
700 ºC. As shown in Fig. 1(c), above 750 ◦ C, the changes began to occur
in the region I and II, circled in black. The vacuum level (3.0 × 10− 5
Torr) did not change, suggesting that dissolution into Ni started to occur.
Reported that the precipitation and segregation reactions have been
observed shown from 725 ◦ C to 900 ◦ C [24]. In addition, In-situ SEM
observations also show dissolution into Ni at a similar temperature
range. [16]
Therefore, it is revealed that the change is due to the carbon dis­
solved in Ni started to re-deposit or the surface MLG started to dissolve
into Ni, during the warming-up process. Afterwards, the observation
with heating as shown in Fig. 1(d) - (f), the phenomenon that occurred in
the image of Fig. 1(c) triggered a gradual change in the appearance of
the entire sample as the temperature increased. From the image in Fig. 1
(e) - (f), the surface change in region III is more pronounced than in the
lower temperature image, and the difference in the change in the three
regions is likely due to the crystal orientation and surface shape of Ni.
[16] Compared to regions I and II, region III is assumed to be a place
with fewer irregularities that are an obstacle to solid solution and pre­
cipitation. There are black lumps on the surface of regions I and II. The
carbon that could not be solidly solubilized in the uneven areas has
become amorphous carbon. It was observed at 900 ◦ C for about 5 min
(See the movie in the video). In addition, the field of view is narrower
than in Fig. 1 due to the specifications of the device. Further observation
with heating changed the state of the surface significantly. From the
movie, it is observed that appears to be "green" on the whole is
decreasing. In the case of graphene with multiple layers, the RGB count
of the CCD can theoretically be shown to have a higher G (green) count
than R (red) and B (blue) [25]..
On the other hand, when the sample was heated in Ar + H2, the dark
blue and orange areas appeared on the sample surface in Fig. 2. This
reacting process was different from that at vacuum condition. These
areas are amorphous carbon (a-C). Half of the graphene area was etched
by the Ar + H2 gas. While keeping the sample at 900 ◦ C in the atmo­
sphere, the graphene was etched away gradually (Fig. 2e~i).
3.2. Observation of graphene of Cu foil
Fig. 3 shows that the change of CVD-grown graphene films on Cu foil
during heated in vacuum and Ar + H2 gas atmosphere by an in-situ OM
M. Kato et al. Applied Surface Science Advances 6 (2021) 100138

Fig. 2. In-situ OM images of CVD-grown graphene on the Ni sample surface during temperature rising in Ar + H2 atmosphere.

Fig. 3. OM images of CVD-grown graphene films on Cu foil during heated in vacuum (a) and Ar + H2 gas atmosphere (b).

3
M. Kato et al.
observation. Fig. 3a hardly shows the color change until the temperature
up to 400 ◦ C during heated in vacuum. With increasing the temperature
to 500 ◦ C, the color change is obvious. While with further heating over
800 ◦ C, the change is slight. The color changes in the OM image might be
caused by the graphene decomposition, which is related on the escape of
intercalated gas, such as residual oxygen between the metal and gra­
phene, and/or the removal of the oxide film [26-28]. While the change
of CVD-grown graphene films on Cu foil during heated in Ar + H2 gas
atmosphere is shown as Fig. 3b. Graphene is stable with lower than
400 ◦ C. With increasing the temperature to 700 ◦ C, the color of graphene
becomes part of the green, which might be caused by the etching away
by the Ar + H2 atmosphere at high temperatures. With further heating
over 900 ◦ C, the change is obvious, which hints that the graphene films
have completely been etched.
It is known that high concentrations of hydrogen and oxygen on the
sample surface can from highly volatile molecules such as CH4 and CO2
[28,29]. From the experiment in 3.1 and 3.2, we could assume that the
chemical reaction between oxygen and hydrogen on the sample surface
or intercalated in the sample layer formed highly volatile molecules and
etched the graphene. The color of the metal surface also changed when
comparing the observation in vacuum and the observation with the
introduction of argon hydrogen. This result indicates that the underlying
metal was also affected in some way [30]. Therefore, in addition to the
basic parameter such as temperature and pressure, impurities formed by
chemical reaction on the sample surface also affect the observation.
3.3. In-situ Raman and XPS analyzed
To further investigate the influence of the atmosphere on heated
graphene on Ni and Cu foil, the change of G-band peak position during
heated in vacuum and Ar + H2 gas atmosphere is summarized in Fig. 4.
Fig. 4 shows the in-situ Raman spectroscopy measured from the CVD-
grown graphene films on Ni and Cu foil during heated in vacuum and
Ar + H2 gas atmosphere. From Fig. 4, the signal of the G-band of gra­
phene films could be found till 900 ºC on Ni foil and till 600 ºC on Cu foil
in vacuum condition. This might be caused by the strong background,
and/or the damage of graphene networks, which is matched with that of
in-situ OM. While heated in Ar + H2 gas atmosphere, Fig. 4 shows that
the signal of G-band disappears at the temperature of 800 ºC by Ni foil
and 600 ºC by Cu foil. Excepting Ni in Vacuum experiment condition,
there is a difference between the in-situ OM image and the disappear­
ance of the signal in the Raman spectra. We thought this was because the
scale of Raman laser irradiation was very different from the scale of in-
Fig. 4. G-band peak position (cm-1) of Raman spectroscopy for graphene films on Ni and Cu foil during heated in vacuum and Ar + H2 gas atmosphere.
4
Applied Surface Science Advances 6 (2021) 100138
situ observation. The information from Raman spectroscopy reflects
information from a much smaller area. It hints that the graphene films
were etching away with H2 during heated in Ar + H2 gas atmosphere at
high temperatures. In general, Raman peaks are temperature dependent,
with the peak position shifting to the lower frequency side and the in­
tensity decreasing as the temperature of the sample increases. The same
trend was obtained in this experiment. The temperature dependence of
the Raman shift in the G-band is independent of the sample. Therefore,
in the range where the G-band can be observed, we can say that the
temperature of graphene on the substrate can be converted from the
Raman shift of the G-band. The slope of the G-band versus temperature
was about − 0.024 cm− 1/ ◦ C for Ni and Cu foils. However, in the case of
heated in Ar + H2 gas atmosphere, the G-band disappeared at 600 ºC,
due to the etching effect of H2 at high temperatures. The graphene films
were completely etched away at 900 ºC, which has been confirmed by
the in-situ OM observation, as shown in Fig. 2 and Fig. 3.
Fig. 5 shows that C 1 s XPS spectroscopy of graphene films on Ni and
Cu foil, before and after heating in vacuum at 900 ºC for 10 min,
respectively. From the C 1 s XPS spectroscopy in Fig. 5, the peaks at
287.5, 286.1 and 284.6 eV can be attributed to the bonds of C=O, C–O,
and C–C, respectively. The presence of the peaks at 287.5 and 286.1 eV
indicates the replacement of C atom by O atom, which might be from
residual oxygen between the metal and graphene during CVD fabrica­
tion [27,31]. It could be seen that the C 1 s intensity peak shifts to higher
energy after heating, which could be caused by metal distortion, and the
escape of intercalated gas such as residual oxygen between the metal
and graphene [26,27,31]. From the XPS results, the peaks 286.1 eV and
287.5 eV, which indicate bonding with oxygen, are less intense when
graphene on nickel is heated compared to graphene on Cu, as shown in
(a) and (b) of Fig. 5. Therefore, the difference in the in-situ optical im­
ages when heated in a vacuum may be because the change on Ni is due to
the "dissolution or precipitation of graphene on the metal surface", while
the change on Cu is due to the "reaction between graphene on the metal
surface and oxygen adsorbed on the surface or intercalated between the
surface and the substrate". The optical images obtained from the ex­
periments are shown in Fig. 1 and Fig. 3(a), the optical images obtained
from the experiments reflect the characteristics of metal catalysts with
different degrees of carbon solidification.
4. Conclusion
In summary, we have investigated the influence of the CVD-grown
graphene films on metal foil during heated in vacuum and Ar + H2
M. Kato et al.
Fig. 5. C 1 s XPS spectroscopy of CVD-grown graphene films on Ni (a) and Cu (b) foil.
gas atmosphere, by combining a portable Raman spectroscopy system
with the in-situ OM observation. The change of MLG on Ni and graphene
on Cu could be recognized by optical microscope images. We have
demonstrated that the in-situ observation of graphene decomposition
has been carried out by using an optical microscope equipped with a
heating vacuum chamber. The characteristics of surface change corre­
sponded to the condition of graphene growth as the using Ni and Cu
catalyst. The G-band peak shifts to lower energy side, with increasing
heated temperature in vacuum and Ar + H2 gas atmosphere, respec­
tively. This method is simple and effective enough to obtain information
on the growth conditions of large areas of graphene on a substrate such
as Ni, which has a high degree of carbon solubility. It could be able to
observe of graphene on the metal easily by optical microscopy and
Raman spectrum could be capable of atmosphere control as well as an
5
Applied Surface Science Advances 6 (2021) 100138
observation by real time. Our results are real time observation with color
image and will provide a simple way of understanding the graphene
state in various atmosphere condition.
Declaration of Competing Interests
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Supplementary materials
Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.apsadv.2021.100138.
M. Kato et al. Applied Surface Science Advances 6 (2021) 100138

References [16] K. Takahashi, K. Yamada, H. Kato, H. Hibino, Y. Homma, Surf. Sci. 606 (7) (2012)
728–732.
[17] Z.J. Wang, G. Weinberg, Q. Zhang, T. Lunkenbein, A. Klein-Hoffmann,
[1] A.K. Geim, K.S. Novoselov, Nature Mater. 6 (2007) 183–191.
M. Kurnatowska, M. Plodinec, Q. Li, L. Chi, R. Schloegl, M.G. Willinger, ACS Nano
[2] F. Bonaccorso, Z. Sun, T.A. Hasan, A.C. Ferrari, Nat. Photonics 4 (9) (2010)
9 (2) (2015) 1506–1519.
611–622.
[18] P. Yan, Y.J. Jeong, M.F. Islam, P.C. Pistorius, RSC Adv. 6 (103) (2016)
[3] A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M.S. Dresselhaus, J. Kong,
101235–101241.
Nano Lett. 9 (1) (2008) 30–35.
[19] C.T. Pan, J.A. Hinks, Q.M. Ramasse, G. Greaves, U. Bangert, S.E. Donnelly, S.
[4] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner, A. Velamakanni, I. Jung,
J. Haigh, Sci. Rep. 4 (1) (2014) 1–7.
E. Tutuc, K.S. Banerjee, L. Colombo, R.S. Ruoff, Science 324 (5932) (2009)
[20] Z. Liu, Y.C. Lin, C.C. Lu, C.H. Yeh, P.W. Chiu, S. Iijima, K. Suenaga, Nat. Commun.
1312–1314.
(2014) 5.
[5] X. Li, C.W. Magnuson, A. Venugopal, R.M. Tromp, J.B. Hannon, E.M. Vogel,
[21] K. Kim, J.C. Yoon, J. Kim, J.H. Kim, S.W. Lee, A. Yoon, Z. Lee, Applied Microscopy
L. Colombo, R.S. Ruoff, J. Am. Chem. Soc. 133 (9) (2011) 2816–2819.
49 (1) (2019) 1–8.
[6] D. Wei, Y. Liu, Y. Wang, H. Zhang, L. Huang, G. Yu, Nano Lett. 9 (5) (2009)
[22] T.O. Terasawa, K. Saiki, Nat. Commun. 6 (2015) 6834.
1752–1758.
[23] T. Taira, S. Obata, K. Saiki, Appl. Phys. Express 10 (5) (2017), 055502.
[7] D.R. Lenski, M.S. Fuhrer, J. Appl. Phys. 110 (1) (2011), 013720.
[24] A. Dahal, M. Batzill, Graphene–nickel interfaces: a review, Nanoscale 6 (5) (2014)
[8] S. Bae, H. Kim, Y. Lee, X. Xu, J.S. Park, Y. Zheng, J. Balakrishnam, T. Lei, R.H. Kim,
2548–2562.
I.Y. Song, J.Y. Kim, S.K. Kim, B. Özyilmaz, J.H. Ahn, H.B. Hong, S. Iijima, Nat.
[25] Y.Y. Wang, R.X. Gao, Z.H. Ni, H. He, S.P. Guo, H.P. Yang, C.X. Cong, T. Yu,
Nanotechnol. 5 (8) (2010) 574–578.
Nanotechnology 23 (49) (2012), 495713.
[9] L. Baraton, Z.B. He, C.S. Lee, C.S. Cojocaru, M. Châtelet, J.L. Maurice, H.Y. Lee,
[26] P. Sutter, J.T. Sadowski, E.A. Sutter, J. Am. Chem. Soc. 132 (2010) 8175–8179.
D. Pribat, EPL (Europhysics Letters) 96 (4) (2011) 46003.
[27] P.R. Kidambi, B.C. Bayer, R. Blume, Z. Wang, C. Baehtz, R.S. Weatherup, M.
[10] F. Abild-Pedersen, J.K. Nørskov, J.R. Rostrup-Nielsen, J. Sehested, S. Helveg, Phys.
G. Willinger, R. Schloegl, S. Hofmann, Nano Lett. 13 (2013) 4769–4778.
Rev. B 73 (11) (2006), 115419.
[28] J. Cho, L. Gao, J. Tian, H. Cao, W. Wu, Q. Yu, E.N. Yitamben, B. Fisher, J.R. Guest,
[11] Y. Wang, A.J. Page, Y. Nishimoto, H.J. Qian, K. Morokuma, S. Irle, J. Am. Chem.
Y. Chen, N.P. Guisinger, ACS Nano 5 (2011) 3607–3613.
Soc. 133 (46) (2011) 18837–18842.
[29] E. Starodub, N.C. Bartelt, K.F. McCarty, J. Phys. Chem. C 114 (11) (2010)
[12] C. Mattevi, H. Kim, M. Chhowalla, J. Mater. Chem. 21 (10) (2011) 3324–3334.
5134–5140.
[13] J.D. Wood, S.W. Schmucker, A.S. Lyons, E. Pop, J.W. Lyding, Nano Lett. 11 (11)
[30] S. Choubak, M. Biron, P.L. Levesque, R. Martel, P. Desjardins, J. Phys. Chem. Lett. 4
(2011) 4547–4554.
(7) (2013) 1100–1103.
[14] X. Li, C.W. Magnuson, A. Venugopal, J. An, J.W. Suk, B. Han, M. Borysiak, W. Cai,
[31] R. Larciprete, S. Ulstrup, P. Lacovig, M. Dalmiglio, M. Bianchi, F. Mazzola,
A. Velamalanni, Y. Zhu, L. Fu, M.E. Vogel, E. Voelkl, L. Colombo, R.S. Ruoff, Nano
L. Hornekær, F. Orlando, A. Baraldi, P. Hofmann, S. Lizzit, ACS Nano 6 (2012)
Lett. 10 (11) (2010) 4328–4334.
9551–9558.
[15] G. Odahara, H. Hibino, N. Nakayama, T. Shimbata, C. Oshima, S. Otani, M. Suzuki,
T. Yasue, T. Koshikawa, Appl. Phys. Express 5 (3) (2012), 035501.

6
 Journal of Non-Crystalline Solids 381 (2013) 12–16

Contents lists available at ScienceDirect

Journal of Non-Crystalline Solids

journal homepage: www.elsevier.com/ locate/ jnoncrysol

Dehydroxilation and crystallization of glasses: A DTA study

D.C. Lago a,⁎, M.O. Prado a,b,c
a
Consejo Nacional de Investigaciones Científicas y Técnicas, Avda. Ezequiel Bustillo Km. 9,5. Código Postal 8400, San Carlos de Bariloche, Pcia. Río Negro, Argentina
b
Departamento Materiales Nucleares, Centro Atómico Bariloche, Comisión Nacional de Energía Atómica, Avda. Ezequiel Bustillo Km. 9,5. Código Postal 8400, San Carlos de Bariloche,
Pcia. Río Negro, Argentina
c
Instituto Balseiro (Universidad Nacional de Cuyo, Comisión Nacional de Energía Atómica), Avda. Ezequiel Bustillo Km. 9,5. Código Postal 8400, San Carlos de Bariloche, Pcia. Río Negro, Argentina

a r t i c l e i n f o a b s t r a c t

Article history: Differential thermal analysis (DTA) is a technique frequently used for the study of the crystallization kinetics of
Received 5 July 2013 glass powders during sintering. The effect of specific surface area on crystallization kinetics has been reported
Received in revised form 31 August 2013 earlier. It has been shown that there is a shift in the DTA crystallization peaks towards lower temperatures for
Available online 2 October 2013
smaller particle size. This paper analyzes how the dehydroxilation endothermic peaks that appear in DTA
measurements depend on specific area and on the heating rate. Experiments were done on yttrium aluminosil-
Keywords:
Dehydroxilation;
icate, samarium aluminosilicate and soda-lime silica glasses, slabs and powders. We found that some of them
Crystallization; exhibit a dehydroxilation endothermic peak that overlaps at different extent with the crystallization peak. The
Yttrium aluminosilicate glass; dehydroxilation peak is not detected when yttrium aluminosilicate glass powder is preheated at melting temper-
Differential Thermal Analysis atures, or its surface is previously hydrated. For the yttrium aluminosilicate glass a dehydroxilation apparent
activation energy of 255 ± 50 kcal mol−1 was obtained using the Kissinger plot.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction
Differential scanning calorimetry (DSC) and differential thermal
analysis (DTA) have been extensively used in the study of the crystalli-
zation kinetics of glasses. Barandiaran and Colmenero [1] used DTA runs
to estimate the critical cooling rates of glasses. Cabral et al. [2] used DTA
to study the stability on heating of glasses that nucleate homogeneous-
ly. Both last studies were run with glass slabs instead of powders in
order to avoid surface crystallization.
There are other papers of DTA measurements, run on glass powders.
One of them is that of R. Müller [3]. Müller presented a model for surface
crystallization kinetics of glasses and compared the model results with
experimental data on fresnoite glass. In Ref. [3], the model satisfactorily
explains the shift towards lower temperatures of the crystallization DTA
peaks for smaller powder particle size. Although not discussed in Ref.
[3], DTA runs on some of the fresnoite glass powders exhibited an endo-
thermic event before the crystallization peak, especially for particles
with sizes of 112 and 65 micrometers. For particles of 1.25 micrometers
no endothermic peak is seen.
In Ref. [4], C. Lara et al. studied the crystallization of RO–BaO–SiO2
(R = Mg, Zn), and it is shown in their DTA runs that for R = Zn there
is an endothermic peak before the crystallization one, but no when
R = Mg.
In Ref. [5], Sung-Bum Sohn et al. studied the influence of CeO2 addi-
tion on the crystallization behavior in the glass system MgO–Al2O3–
⁎ Corresponding author at: Centro Atómico Bariloche, Avenida Bustillo 9500, R8402AGP,
S.C. de Bariloche, Río Negro, Argentina. Tel./fax: +54 294 4445241.
E-mail address: dlago@cab.cnea.gov.ar (D.C. Lago).
SiO2, and also showed an endothermic peak before the crystallization
peak.
In Ref. [6], Reben M., in a thermal study of oxyfluoride glass with
strontium fluoride, also found an endothermic peak before the crystalli-
zation exothermic event.
This endothermic peak before crystallization is generally not
discussed in the glass literature although it is thoroughly studied in
mineral structure works during heating [7,8]. Moreover, very often the
temperature for crystallization start, TS, is chosen at the minimum of
the endothermic peak which we will show that is not correct.
2. Experimental
2.1. Glass preparation
Yttrium aluminosilicate glass samples were prepared with composi-
tion: 1Y2O3 · 1Al2O3 · 3SiO2 which corresponds approximately to the
core of the ternary glassy domain [9]. Samarium aluminosilicate glass
samples were prepared with composition: 1Sm2O3 · 1Al2O3 · 3SiO2.
In each case, a mixture of high purity oxide precursors was homog-
enized in a rotating mill and melted in a platinum crucible for 2 h at
1600 °C. The melt was poured on a stainless steel plate and immediately
splat-cooled using a second steel plate. A fraction of the glass was sepa-
rated and identified as slab. The remainder glass was milled and sieved
to obtain three different powder size samples.
Soda-lime silica glass samples were prepared with composition
1Na2O · 1CaO · 2SiO2 (NCS). The melting procedure was the same as
detailed above, at a lower temperature (1500 °C).

0022-3093/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jnoncrysol.2013.09.008
 D.C. Lago, M.O. Prado / Journal of Non-Crystalline Solids 381 (2013) 12–16
2.2. Particle size distribution measurement
Glass pieces were milled in an RS 200 vibratory disk mill and sieved.
In the case of the yttrium aluminosilicate (YAS), the glass fractions
obtained were those remaining between the following pairs of sieves:
50 and 100 μm (fraction 1), between 25 and 37 μm (fraction 2), and
less than 25 μm (fraction 3). The last particle size was obtained by wet
sieving.
Samarium aluminosilicate glass and soda-lime silica glass particles
between 25 and 37 μm sieves were collected.
The particle size distribution of the different glass fractions was mea-
sured by laser diffraction with a Mastersizer Micro Malvern Instrument,
with simultaneously applied ultrasound to disaggregate particle clus-
ters. The dispersive medium was distilled water.
2.3. Flame polished surface
Part of fraction 2 glass powder was spheroidized by passing by a
propane–oxygen flame. Particle surface got flame polished.
2.4. Thermal and gravimetric simultaneous analysis
The sample weight and the DTA signal were simultaneously mea-
sured with a SDT-Q600 TA instrument.
Glass transition (Tg) and crystallization temperatures (TX) were also
determined. Measurements were done in alumina or platinum cruci-
bles. The reference crucible was an empty crucible. Experiments were
carried out up to 1300 °C with a heating rate of 10 °C min−1, in a
100 ml min−1 flux of 5.0 purity nitrogen gas.
2.5. XRD sample characterization
Crystallinity analysis was carried out on powder samples with a
Bruker D8 Advance diffractometer, at a wavelength of 0.154054 nm.
Scans were performed in the range 10°–90° with a step size of
0.01° min−1.
3. Results
3.1. Particle size distribution
Fig. 1 shows the particle size distribution for each of the YAS glass
powder fractions. As acicular particles pass through the sieves, the mea-
sured particle ranges are broader than the expected distributions.
Fig. 1. Particle size distribution of different YAS glass powder fractions.
13
3.2. Differential thermal analysis
3.2.1. The effect of particle size distribution
Fig. 2 shows the DTA results for a monolith and the different frac-
tions of YAS glass powder in alumina crucibles. The trace corresponding
to the slab exhibits the Tg inflection and the crystallization peak as
expected (being TS the temperature corresponding to the start of the
crystallization peak and TX, the temperature of the peak maximum).
The traces corresponding to glass powders present the same structure
as for the slab, but moreover they have an endothermic peak at TE.
The results of the DTA analysis are resumed in Table 1.
Fig. 3 shows the overlapping of the DTA trace for YAS fraction 3 and
the cross section area of a powder cylinder during sintering in a hot
stage microscope (HSM). Shrinking is previous to the endothermic
peak.
3.2.2. The effect of annealing near Tg
In Fig. 4, we show the thermal behavior of YAS glass powder (frac-
tion 2) as in Fig. 2, with legend 1a and 3a the effect of adding annealing
1 h and 3 h at 900 °C respectively. This was done in order to allow
the system to stay some time in a viscoelastic region, during a character-
istic time t = η/G. η = 1013 Pa s and G is the shear modulus with an
approximated value of 1010 Pa. This gives 1000 s as a characteristic re-
laxation time.
3.2.3. The effect of crucible chemical composition
In order to confirm that there was no chemical reaction between the
glass powder sample and the alumina crucible, we repeated the mea-
surement with particles between 25 and 37 μm (fraction 2) but this
time in a platinum crucible (the results are shown in Fig. 4).
In order to discard the possible formation of crystalline phases dur-
ing the endothermic peak, in one of the experiments the thermal
heating was stopped at 1040 °C (just at the maximum of the endother-
mic peak), and the sample was quenched to room temperature at
50 °C min−1. XRD was performed on the resulting sample.
3.2.4. The effect of particle surface quality
Since the DTA exothermic peak is observed with YAS glass powder,
but not with the slab, the physical origin of this peak must be linked
to glass surface properties. In order to explore this fact, in Fig. 5 we com-
pare the DTA traces for YAS glass powder fraction 2, YAS glass powder
fraction 2 + 30 h in distilled water at room temperature and YAS
glass powder spheroidized in propane–oxygen flame.
The endothermic peak is well defined for the powder, but only sug-
gested for powder that was in contact with water or that was passed
through a propane–oxygen flame.
Fig. 2. DTA traces for slab and different YAS glass powder fractions. Tg, TE, TS and TX are
indicated.
14 D.C. Lago, M.O. Prado / Journal of Non-Crystalline Solids 381 (2013) 12–16

Table 1
Glass transition and crystallization temperatures for different YAS fractions.

Sample Tg (°C) TE (°C) TS (°C) TX (°C)

Slab 920 – 1148 1243

Fraction 1 919 1062 1057 1153
Fraction 2 917 1044 1037 1126
Fraction 3 922 1037 1033 1122

3.2.5. Other glasses

3.2.5.1. Samarium aluminosilicate glass. In this paper we also studied the

DTA behavior of other glasses. Samarium aluminosilicate glass slab ex-
hibits the same behavior as yttrium aluminosilicate glass slab; it does
not present an endothermic peak previous to the crystallization peak.
However, samarium aluminosilicate glass powder does (see Fig. 6), Fig. 4. 0a: thermal behavior of YAS glass powder (fraction 2). 1a: the same measurement,
confirming the fact that the endothermic peak is a physical event linked but adding an isothermal plateau at Tg = 900 °C during 1 h. 3a: isothermal plateau at
to the specific surface area. 900 °C during 3 h. Pt: thermal behavior of YAS glass powder (fraction 2) in platinum
crucible.

3.2.5.2. Soda-lime silica glass. A soda-lime metasilicate glass (NCS) was
also tested. Fig. 7 shows the DTA trace for a slab and powder of a NCS
glass. No endothermic peak is observed.
3.2.6. The effect of the heating rate on the endothermic reaction
For the YAS glass powder, fraction 2, DTA runs were done at 5, 10, 15,
20 and 40 °C min−1 (see Fig. 8). A Kissinger plot of the temperatures of
maximum endothermic reaction rate is shown in Fig. 9. An apparent
activation energy of 255 ± 50 kcal mol−1 is determined for the endo-
thermic reaction.
3.2.7. Thermogravimetry
Fig. 10 shows thermogravimetric behavior of a YAS slab and of YAS
powder with three different surfaces.
Let us remark that only YAS powder (b) exhibits endothermic peak
in Fig. 5.
We assign the initial mass increase of the slab (a) to measurement
error. Above Tg there is a mass decrease of about 1o/oo that we assign
to a decrease in dissolved gases. For temperatures higher than TS there
is a slight mass decrease due to the decrease of gases solubility in the
crystalline phases.
Glass microspheres (d) show at about 300 °C a slight mass decrease
due to physical adsorbed water and again above Tg due to a decrease in
dissolved gases. Both processes show very slight mass changes since
during flame spheroidization most water and dissolved gases were
expulsed.
YAS powder (b) and YAS powder 30 h in water (c) show at about
300 °C a loss of adsorbed water, chemically adsorbed water at 600 °C
and gas expulsion due to crystallization above 1122 °C.
From the results shown in this section, we do not detect a change
of weight during the endothermic event. A simple calculation using
Zhuravlev (Ref. [13]) data of maximum silanols (≡Si―OH) surface den-
sity equal to 5 ≡SiOH per nm−2 in silica, and considering that our com-
position is 1Y2O3 · 1Al2O3 · 3SiO2 and that 10% of surface Si has a
silanol functional group, 3000 cm2∙gr−1 sample specific surface area,
and sample mass approximately 50 mg, shows that the mass loss dur-
ing dehydration should be less than 0.3 micrograms, i.e. near the sensi-
bility of our TG equipment.
4. Discussion
Fig. 2 shows a shift of the crystallization peaks of YAS powder sam-
ples towards lower temperatures as expected from Ref. [3]. However,
the endothermic peaks exhibited by all the YAS powder samples except
by the monolithic one are not explained in the literature.
The same occurs with SmAS glasses and other observed in this paper
and the referenced literature.
The glass transition temperature is the same for all YAS samples,
being in slab or powder form, suggesting that there has not been intro-
duced glass structure changes during grinding. If grinding changed the
glass structure by plastic deformation, this plastically deformed glass
should have a different Tg value, and this is not observed. The glass tran-
sition temperature is maintained in all YAS samples, indicating that the
supercooled liquid structure inherited is the same for all samples. We

Fig. 3. Comparison of DTA and HSM curves on the same temperature scale for YAS glass Fig. 5. DTA of YAS glass: (a) slab, (b) powder (fraction 2), (c) powder (fraction 2) + 30 h
powder (heating rate 10 °C min−1). Solid line: DTA trace; dot line: HSM trace. in contact with water at room temperature, (d) glass powder microspheres.
 D.C. Lago, M.O. Prado / Journal of Non-Crystalline Solids 381 (2013) 12–16
Fig. 6. DTA of (a) YAS glass slab, (b) SmAS glass slab and (c) SmAS glass powder showing
the endothermic peak before crystallization.
note that the energy involved in this endothermic peak is comparable to
the crystallization enthalpy.
The endothermic peak is due to a surface effect during heating of
the crushed sample, since the monolithic sample does not present the
endothermic behavior.
However, neither a soda-lime silicate glass powder nor a Pyrex glass
powder tested in this work (DTA not shown in this work) exhibits endo-
thermic peak (Fig. 7).
Moreover in the literature, some glass powder compositions exhibit
a DTA endothermic peak [3-5] and some do not [10,11]. In Ref. [4], mea-
surements on two sintering glass powders are shown; one of them ex-
hibits endothermic peak and the other does not.
The partial overlapping of this peak with the crystallization peak
complicates the interpretation of this evidence, which is more clear in
the case of the YAS glass powder fraction 3 where the endothermic
peak is well defined before the crystallization peak.
With YAS glass powders, the endo-peak is observable either using
alumina crucibles or platinum crucibles, which eliminates the possibili-
ty that it is due to a sample-alumina crucible chemical reaction.
Moreover, the X-ray diffraction analysis confirms that within the
limit of detection of XRD there is no crystallization at 1040 °C endother-
mic peak.
Fig. 5 shows that the endo-peak becomes sluggish when the YAS
glass powder is passed through a propane–oxygen flame or when the
powder is put for about 30 h in contact with distilled water.
It was shown for clays [7] that a possible reaction mechanism for
dehydroxilation can be proposed. The proposed mechanism needs a
proton delocalization at a specific hydroxyl site. For H2O to form, the
delocalized proton must migrate to a second hydroxyl site. If however,
Fig. 7. DTA of (a) NCS glass slab, and (b) NCS glass powder showing the absence of an en-
dothermic peak before crystallization.
15
Fig. 8. Heating rate effect on the DTA results for YAS glass powder (fraction 2). The endo-
thermic and the crystallization peak shifts towards higher temperatures with the increase
of the heating rate.
nonadjacent hydroxyls are involved, proton diffusion is needed, as
was proved by the electrical conductivity increase in measurements
done up to the dehydroxylation temperature.
We expect that the particles that were flame polished contain a
lower surface density of silanols (≡Si―OH) which conducts to the
almost absence of the endothermic peak (Fig. 5d). In the case of glass
powder previously in contact with water (Fig. 5c), the increase of sur-
face silanol density allows the condensation reaction [12,13]:
≡Si―OH þ HO―Si≡→≡Si―O―Si≡ þ H2 O↑
to take place at lower temperatures, showing an important mass loss
below 600 °C (Fig. 10c).
Thus, this surface endothermic reaction, with an apparent reaction
enthalpy of 255 kcal mol−1 for YAS glass, as shown in Fig. 9, can be
assigned to a dehydroxilation reaction. This value is in agreement with
dehydroxilation measurements for clays [7].
5. Conclusions
The correct interpretation of the DTA/DSC experiments spectra is
possible as long as we are aware of the different physical phenomena
that occur during glass powder heating. Particularly, on studying glass
powders crystallization, there is an effect on the crystallization peaks
of moving toward lower temperatures, when compared with the results
obtained with glass slabs.
Moreover when run on some glass powders, YAS glasses for exam-
ple, DTA traces exhibit an endothermic peak before (or overlapped
Fig. 9. Kissinger plot of YAS temperatures of maximum crystallization rates at different
heating rates (data from Fig. 7).
16 D.C. Lago, M.O. Prado / Journal of Non-Crystalline Solids 381 (2013) 12–16
Fig. 10. Thermogravimetry of YAS samples with different surface quality.
with) the crystallization peak. This endothermic peak appears when we
work with glass powders indicating that it is due to an endothermic sur-
face process.
This endothermic peak has height, width and temperature span that
depends upon the glass surface chemistry.
As the process is endothermic, it is not due to strain relaxation, crys-
tallization (we checked experimentally this point with XRD just at the
maximum of the endothermic peak) or available energy due to the
sintering decreasing surface. Sintering can show a change in the DTA
baseline but not an endothermic peak.
Since YAS glass components have a unique oxidation state, we do
not expect an endochemical (or even exothermic) reaction to take
place during the heating process.
The fact that flame spheroidized particles dramatically decrease the
endothermic peak area suggests that the lower surface density of
silanols results in a lower reaction heat consumption.
We determined that the apparent activation energy for this endo-
thermic process is about 255 kcal mol−1, which is in agreement with
the dehydroxilation energy in clays.
In order to analyze a glass powder DTA, it is convenient to compare it
with a slab DTA of the same glass, since in the latter surface effects are
minimized.
Acknowledgements
This work has been financed by the project PICT 2008-39 ANPCyT
Argentina.
References
[1] J.M. Barandiaran, J. Colmenero, J. Non-Cryst. Solids 46 (1981) 277–287.
[2] A. Cabral, A. Cardoso, E.D. Zanotto, J. Non-Cryst. Solids 320 (2003) 1–8.
[3] R. Müller, J. Therm. Anal. 35 (1989) 823–835.
[4] C. Lara, M.J. Pascual, A. Durán, J. Non-Cryst. Solids 348 (2004) 149–155.
[5] Sung-Bum Sohn, Se-Young Choi, J. Non-Cryst. Solids 282 (2001) 221–227.
[6] Manuela Reben, J. Non-Cryst. Solids 357 (2011) 2653–2657.
[7] R.L. Frost, A.M. Vasallo, Clay Clay Miner. 44 (5) (1996) 635–651.
[8] R.E. Grim, W.F. Bradley, Am. Mineral. 33 (1–2) (1948) 50–59.
[9] J.E. White, D.E. Day, Key Eng. Mater. 181 (1994) 94–95.
[10] Xiaojie J. Xu, Chandra S. Ray, Delbert E. Day, J. Am. Ceram. Soc. 74 (5) (1991)
909–914.
[11] Najim Sadiki, Jean Pierre Coutures, Catherine Fillet, Jean Luc Dussossoy, J. Nucl.
Mater. 348 (2006) 70–78.
[12] Quan Wan, Christopher Ramsey, George Baran, J. Therm. Anal. Calorim. 99–1 (2010)
237–243.
[13] L.T. Zhuravlev, Colloids Surf. A Physicochem. Eng. Asp. 173 (2000) 1–38.
 Radiation Physics and Chemistry 181 (2021) 109334

Contents lists available at ScienceDirect

Radiation Physics and Chemistry

journal homepage: www.elsevier.com/locate/radphyschem

Femtosecond laser-induced X-ray fluorescence measurement of

metal-protein compounds
Jungu Kang, Do-Kyeong Ko *
Department of Physics and Photon Science, Gwangju Institute of Science and Technology, Gwangju, 61005, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: We measured the metal-protein compounds X-ray fluorescence by the laser-induced X-ray source with an
Laser plasma aluminum tape target. The laser-induced X-ray generated by the high repetition rate (1 kHz) femtosecond (40 fs)
High repetition rate Ti: Sapphire laser system. To characterize a laser-induced Kα and bremsstrahlung X-ray source, we changed the
Kα
laser intensity and pulse focused position. We measured the X-ray fluorescence signal with the copper plate,
Laser-induced X-ray
X-ray fluorescence
selenium powder, lyophilized hemoglobin, and lyophilized seleno-methionine, which was detected by X-ray
Metal-protein spectrometer. With a few micrograms of samples, the X-ray fluorescence can be measured with high accuracy and
high stability.

1. Introduction parameters can be optimized for each specific experimental purpose.

Since 1895 when X-rays were first discovered, various types of X-ray
sources have been developed they have been used as important tools in
many fields due to its short wavelength, high energy, and the ability to
penetrate substance. Among the application researches using the X-ray,
the micro X-ray fluorescence has been actively researched recently due
to the advantage of having a high spatial resolution. Usually the micro X-
ray fluorescence experiments with the micro X-ray tube performed with
the poly capillary optics. The micro x-ray source is focused by using the
poly capillary and has a relatively higher flux than the conventional
micro X-ray source. The transmission efficiency of the poly capillary is
affected by the focusing position and the X-ray energy. Another micro­
focus X-ray source for micro X-ray fluorescence is a liquid metal jet X-ray
source that uses liquid metal as an anode. The liquid metal jet X-ray
source has the advantage of being able to adjust the source size on a
micro-scale and having a relatively high flux at the laboratory scale.
However, since there is a lot of debris emissions and the X-ray emission
spot is relatively unstable, various research groups are researching to
resolve debris emission and instability (Otendal et al., 2007). The syn­
chrotron X-ray source has high flux, but the synchrotron source is quite
big, quite expensive, and is not easy to access.
The laser-induced X-ray sources are cheaper than huge synchrotron
X-ray sources and could be built in table-top scale. Moreover, by
changing the laser pulse parameters such as spectrum, duration, repe­
tition rate, flux, size, and the target materials, the X-ray output
Therefore, the laser-induced X-ray source would be a versatile X-ray
source (Sokolowski-Tinten and von der Linde, 2004; Yu et al., 1999;
Feurer et al., 2001; Li et al., 2017a; Zamponi et al., 2009). And the
laser-induced micro X-ray source does not require poly capillary optics
and has a relatively stable X-ray emission spot. Also, the photon flux is
higher than the micro X-ray tube source. Therefore, the laser-induced
point light X-ray source can be a substitute for microfocus X-ray tubes
in high spatial resolution imaging experiments requiring a small size
X-ray source. The laser-induced X-ray is cost-effective compared to the
synchrotron X-ray source and it has a higher flux compared to the
microfocus X-ray tubes (Iqbal et al., 2014). Therefore, the laser-induced
X-ray can be a good candidate for an X-ray source in between the X-ray
tube and the synchrotron X-ray source.
The laser-induced X-ray sources also have been extensively studied
due to their advantages, such as flexibility, high brightness, and point-
like source (Tillman et al., 1995; H€ofer et al., 2016; Li et al., 2017b;
Larsson et al., 2002; Morak et al., 2006). This short and bright point light
X-ray source can be used for atomic-level structure change observation
in the field of material science. Besides, it is applicable to X-ray ab­
sorption spectroscopy, diffraction, radiography, microscopy, and phase
difference imaging. And recently, the laser-induced X-ray fluorescence
experiments were studied in the fields of artwork and cultural heritage,
where the main topic was the combination and control of Kα and elec­
trons (ValleBrozas et al., 1007; Barberio et al., 1038). This shows that
the laser-induced X-ray source can broaden the application area due to

* Corresponding author.
E-mail address: dkko@gist.ac.kr (D.-K. Ko).

https://doi.org/10.1016/j.radphyschem.2020.109334
Received 20 July 2020; Received in revised form 14 December 2020; Accepted 25 December 2020
Available online 28 December 2020
0969-806X/© 2020 Elsevier Ltd. All rights reserved.
J. Kang and D.-K. Ko Radiation Physics and Chemistry 181 (2021) 109334

its unique characteristics. Various research groups are working on
improving the conversion efficiency and the photon flux of the
laser-induced Kα X-ray by changing laser pulse parameters such as the
incident angle and contrast ratio. (Goryaev et al., 2014; Azamoum et al.,
2018).
We have optimized the characteristic emission of laser-induced X-ray
and also optimized the bremsstrahlung emission required for x-ray
fluorescence measurements. Various types of metal targets such as wire,
plate, tape, rotating wheel, and liquid metal are used to generate laser-
induced X-rays. Tape target is useful for experiments using high repe­
tition rate laser pulses because it allows continuous experiments, but its
stability is relatively low due to laser pulse pressure (Zhang et al., 2014).
To maintain the high stability and accuracy of the laser-induced X-ray
fluorescence measurement, we adjusted the thickness of the tape target
and manufactured a new tape rotating holder. In this study, we devel­
oped the laser-induced X-ray source for XRF by the combination and
control of characteristic emission and bremsstrahlung emission and
measured the fluorescence of metallic substances, metal compounds,
and metal-protein materials (seleno-peptides) with high stability and
high accuracy.
2. Experimental setup and methods
Fig. 1 shows the experimental setup for laser-induced X-ray gener­
ation and fluorescence measurement. This experiment was performed
with a double-amplified CPA femtosecond Ti:sapphire laser source (36 fs
pulses, 805 nm wavelength, 1 kHz repetition rate, 10 W power). All
experiments were performed in a high vacuum environment below 10− 4
Torr with a vacuum chamber. Various types of solid targets such as wire,
tape, wheel, and plate were used, depending on the characteristics of the
laser used for X-ray generation. We used a 1 kHz high repetition rate
laser, so we used a tape target for continuous X-ray generation. The
focused high energy pulse creates a micrometer size crater on the target.
Since generation efficiency is lowered due to the crater, the laser must be
irradiated with pulse-by-pulse in the non-damaged area (Goryaev et al.,
2014). We rotated the tape target at a constant rate of 4 cm/s consid­
ering the crater size and repetition rate. The thickness of the tape target
was 40 μm, which was designed considering the metal strength and the
fluctuation due to the focusing of the laser pulse. The disadvantages of
using a tape target are frequent fluctuations and low stability (Zhang
et al., 2014). Therefore, we designed and manufactured the tape rotation
mount to improve the stability of X-ray generation. The Al target has low
threshold energy and can be generated with high efficiency, even with

Fig. 1. Top view of the experimental setup for laser-induced X-ray generation with a femtosecond laser. The experimental setup consists of two parts, (a) opti­
mization of laser-induced X-ray source by controlling the laser intensity, focal position, and pulse width and (b) measurement of X-ray fluorescence by the laser-
induced source. All experiments were performed in a vacuum chamber of 10-4 Torr or less

2
J. Kang and D.-K. Ko
relatively low laser intensity. Since this property is suitable for compact
table-top X-ray sources, we used an Al target. In addition, Al has the
advantages of low cost, easy fabrication, and disposal.
The femtosecond laser pulses focused at an intensity of 1016 W/cm2
or more by the off-axis parabolic mirror. The laser-induced X-ray gen­
eration efficiency varies with the angle at which the focused laser irra­
diates the aluminum tape target. To maximize efficiency, the X-ray was
generated by the laser incident at an angle of 30◦ (Goryaev et al., 2014).
The femtosecond laser-induced X-ray sources have a flux of 107 photo­
n/s⋅srad in the Al Kα energy region and 107 photon/s⋅srad in the
Bremsstrahlung region. Characterizing a laser-induced Kα and brems­
strahlung X-ray source by changed the laser intensity, pulse focused
position, and pulse width (Fig. 1 a). The metal compound X-ray fluo­
rescence measurement is performed by the laser-induced X-ray source,
which has a wide energy range, from 0.1 keV to 20 keV. The metal
compound sample was placed behind the laser-induced X-ray source,
and the metal fluorescence was measured by using an Amptek X-ray
detector (XR-100CR). To reduce noise, the Pb plate was wrapped around
the X-ray detector to block the laser-induced X-ray source (main beam)
and optimized the fluorescence signal by adjusting the position and
velocity of the Al tape target (Fig. 1 b). In order to prevent the noise
caused by the reflected laser pulse and the electrons from the
laser-produced plasma, the X-ray signal was optimized by adjusting the
threshold and the detecting range on the detector program. Using the
magnets to remove electron noise is also a good option, but we chose the
threshold condition control option due to the limitation of the experi­
mental space. To prevent debris, a debris-proof box and acrylic tape
products were manufactured and installed near the pulse focusing site.
Fig. 2. (a) Spectrum of laser-induced X-ray with aluminum. It corresponds to the laser intensity of 1016 W/cm2 and an angle of incidence of 30◦ . (b) Dependence of
the aluminum Kα flux on the femtosecond laser power. (c) Dependence of the aluminum Kα (1.486 keV) and bremsstrahlung emission (>3.5 keV) photon numbers on
the femtosecond laser power. (d) The conversion efficiency of laser-induced aluminum Kα change by the laser power.
3
Radiation Physics and Chemistry 181 (2021) 109334
The laser-induced X-ray fluorescence of copper plate (Cu), human he­
moglobin (Fe), selenium powder (Se), and seleno-peptide (Se) was
measured to research the X-ray fluorescence in various energy ranges.
Metal compound samples were commercially available products (Sigma
Aldrich).
3. Results and discussion
3.1. Al tape target laser-induced X-ray generation and characterization
To improve and characterize the flux of the laser-induced X-ray
source, we controlled a few physical parameters. Fig. 2(a) shows the
spectrum of femtosecond laser-induced X-ray sources with an Al tape
target. This spectrum is the raw spectrum data without concerned the
detector efficiency and any affected with the filter. It corresponds to the
laser power of 1.5 W, the focused beam diameter was 22 μm, and the
peak power of 1.97ⅹ1016 W/cm2. The spectrum shows the Al Kα lines at
1.486 keV and the pile-up peak at 2.972 keV. In the energy region of 3.5
keV or more, the bremsstrahlung continuum emission can be confirmed.
The maximum photon flux of the Kα X-ray source is 4.84ⅹ107 photon/
s∙srad and the photon flux of bremsstrahlung emission is 1.34ⅹ107
photon/s∙srad which is 3.6 times less than the Kα emission. The pile-up
peak at 2.972 keV and 4.5 keV were not included in the line emission
flux calculation.
The metal inner ionization energy threshold depends on the Kα value
of the metal target and laser irradiance value. The ionization threshold
condition is given by (Gibbon and Förster, 1996).
J. Kang and D.-K. Ko
v
√̅̅̅̅̅̅̅̅̅̅̅
2U 1
(
I λ2
)1/2
= = 0.84 / (1)
c me c 1018 W cm2 μm2
where U, I and λ are the Kα characteristic emission energy of the target
material, the laser peak power and the wavelength of the laser pulse. As
the atomic number Z of the target material increased, Kα emission en­
ergy increased, and a higher ionization threshold peak power is
required. According to the calculation, the ionization threshold intensity
of aluminum, copper and molybdenum are 1.28ⅹ1016 W/cm2,
6.92ⅹ1016 W/cm2 and 1.5ⅹ1017 W/cm2, respectively. The generated
electrons are accelerated by the strong laser field and it collides to the
metal. Characteristic emission occurs due to the collision of the accel­
erated electrons and metal inner shell. The laser-induced X-ray source is
affected by the hot electron temperature, and the hot electron temper­
ature is proportional to (Iλ2)α, where α is between 1/3 and 1 depends on
the numerical calculation assumptions of the laser absorption. (Four­
maux and Kieffer, 2016).
In this study, since the aluminum tape target with relatively low Kα
energy was used, the laser-induced metal inner ionization was suffi­
ciently performed, even at a peak power of 1016 W/cm2. Fig. 2(b) is a
peak signal of aluminum Kα according to the change of the laser power.
The Al Kα signal increased with increasing laser power and then satu­
rated from 0.9 W or higher. At the maximum value, the FWHM of the Al
Kα peak is 0.117 keV, and the flux is 5.42ⅹ107 photon/s∙srad. The
dependence of the aluminum Kα (1.486 keV) and the bremsstrahlung
emission (>3.5 keV) photon numbers per second on the femtosecond
laser power is shown in Fig. 2(c). As the laser power increased, the Kα
photon increased linearly, and then it saturated from 0.8 W or higher.
The metal inner ionization threshold power value is 0.7 W, and the
corresponding experimental result satisfies the ionization threshold
condition. Two fitting lines correspond to a Boltzmann fit and the bar is a
standard error with 10 measurement trials. The intensity of the brems­
strahlung emission increased dramatically in the Al Kα saturation re­
gion, after which it also become saturated. The intensity of the
bremsstrahlung emission generated in the laser-induced plasma varies
with the bulk electron temperature. It is described as follows (Alkhi­
mova et al., 2018).
/( )
Ibrs ̃Aexp− 12.4/λTe λ2 (2)
where Ibrs, Te, λ and A are the intensity of the bremsstrahlung emission,
the bulk electron temperature (keV), the wavelength (Å) and a constant.
As the bulk electron temperature increased, the intensity of the Brems­
strahlung emission increased rapidly and become saturated and
converged. The bulk electron temperature can be adjusted by the laser
intensity and focused pulse alignment. (Alkhimova et al., 2018) We
optimized Te and Ibrs, by controlling the corresponding parameters.
Fig. 2(d) shows the conversion efficiency of Al Kα versus the laser
power. The correlation between laser power and conversion efficiency
has already been reported in previous research when using the femto­
second Ti sapphire laser pulse and aluminum target (Nakano et al.,
2004). The optimum laser intensity for conversion efficiency is driven as
(Nakano et al., 2004).
2 with selenium powder is shown in Fig. 5. Fig. 5(b) shows the photon
(Z/13)
(opt)
I16 ≈ 2/3
(3)
(L/λ) sin2 θ + 0.1ηa
where Z, L, θ and ηa are the atomic number of target, the density scale
length, the incident angle of the laser and the absorption coefficient,
respectively. For example, the atomic number of aluminum target Z is
opt
13, the optimal intensity is I16 ≈ 1.5 and, for a copper target (Z = 29),
opt
I16 ≈ 8. (Nakano et al., 2004) According to the theoretical calculation,
when the laser intensity is 1.5ⅹ1016 W/cm2, it has a maximum conver­
sion efficiency value. When the laser power is 1.15 W, the laser peak
power is 1.5ⅹ1016 W/cm2, and it has maximum conversion efficiency. As
shown in Fig. 2(d), when the laser power is 0.9 W, the laser peak power
4
Radiation Physics and Chemistry 181 (2021) 109334
is 1.18ⅹ1016 W/cm2, and the conversion efficiency is 7.86ⅹ10− 6. Two
fitting lines correspond to a lognormal fitting of the experimental points,
which is a guide for the eye and follows the theoretical model. The
corresponding conversion efficiency value is about 80% compared to the
theoretical calculation maximum value of 10− 5.
Fig. 3 shows the dependence of aluminum Kα flux on the laser pulse
width, which changed from 40 fs to 120 fs. Since the laser pulse width
affects the Kα flux, we used the shortest pulse width (36 fs) for all ex­
periments. Fig. 4 shows the Al Kα (1.478 keV) and bremsstrahlung
emission (>3.5 keV) on the change of focal position in the micrometer
scale. We set the focal position zero at the maximum K alpha emission.
Since the scale of the laser-produced plasma is tens of μm, Al Kα and
bremsstrahlung emission were measured 30 times at 10 μm intervals
before and after the focal point. As shown in Fig. 4, the aluminum Kα and
bremsstrahlung emission decreased as the position moved away from
the focal point. As the position of the tape target moved away from the
focal point, the radius of the focused laser increased and the peak power
decreased. Since the peak power change affected the hot electrons
accelerated by the laser field, the Al Kα emission decreased as the Al tape
target moved away from the focal point. The position of the maximum
bremsstrahlung emission is shifted by 10–20 μm relative to the Al Kα,
and the bremsstrahlung emission also decreased as the tape target
moved away from the focal position. The increase of the bulk electron
temperature affects the increase of the bremsstrahlung emission, but it
has the opposite effect on the Kα characteristic emission. According to
previous research, the Ti Kα decreased with the increasing of the bulk
electron temperature (Cristoforetti et al., 2013). When the bulk electron
temperature increased, the Al Kα emission deteriorated, which causes
the Al Kα maximum position to shift by 10–20 μm. At the X-ray fluo­
rescence, the fluorescence signal was optimized at this shifted position.
We could obtain the high-efficiency, compact laser-induced X-ray source
with 80% efficiency of the theoretical calculations and studied the
characteristics on the laser intensity, pulse width and focused position.
3.2. Femtosecond laser-induced X-ray fluorescence measurement of
metal-protein compounds
To precisely calibrate the X-ray fluorescence signal, we measured the
laser-induced X-ray fluorescence with selenium powder and the copper
plate. As shown in Fig. 4, we set the focal position zero at the maximum
K alpha emission and the maximum Bremsstrahlung was obtained at the
position shifted by 10–20 μm from zero position. X-ray fluorescence
measurement was performed at the shifted position. In this experiment,
the X-ray fluorescence produced directly by laser-produced plasma
electrons was not considered. Application experiments are possible
using the penetration depth of the electrons generated in laser produced
plasma (ValleBrozas et al., 1007). Further research can be performed
using the electrons generated by the laser produced plasma.
Fig. 5(a) shows the X-ray fluorescence of the selenium powder and
copper plate. We confirmed the selenium Kα and Kβ signal at 11.22 keV
and 12.49 keV, respectively. The ratio of the Kβ/Kα is 0.159. It shows the
copper Kα and Kβ signal at 8.04 keV and 8.90 keV, respectively. The
ratio of Kβ/Kα is 0.097. The stability of laser-induced X-ray fluorescence
number per second at the selenium Kα peak. The average is 37.75, the
standard deviation is 1.29, and the standard error is 0.41. Fig. 5(b) is the
energy value graph of the X-ray fluorescence of the selenium Kα peak.
The average energy, which was measured 10 times, was 11.218 keV,
with a standard deviation of 0.002 and a standard error of 0.0007. This
data is within the error range of 0.037% compared to the theoretical
value of Selenium Kα 1 energy of 11.2224 keV.
Factors that affect the stability and accuracy of laser-induced X-ray
fluorescence include external pressure, due to the laser pulse focusing. It
causes the change of the flux and peak signal of x-ray fluorescence,
resulting in lower stability and lower accuracy. Therefore, we optimized
the thickness of the Al tape as 40 μm for the experiment and
J. Kang and D.-K. Ko
Fig. 3. Dependence of aluminum Kα flux on the laser pulse width (fs)
Fig. 4. Dependence of aluminum Kα and bremsstrahlung photons on the po­
sition (μm)
manufactured the tape target mount to minimize the fluctuation of the
X-ray fluorescence signal. By doing this, we obtained stable and accurate
laser-induced X-ray fluorescence signal with errors within 0.037%.
The X-ray fluorescence of metal-protein compounds, such as lyoph­
ilized hemoglobin and lyophilized seleno-methionine, was measured
using the laser-induced X-ray light source. To increase the intensity of
the X-ray fluorescence signal, the distance between the sample and the
X-ray detector was compactly reduced, and a Pb blocker was installed to
block the bremsstrahlung noise signal.
Fig. 6 shows the laser-induced X-ray fluorescence spectrum of
lyophilized hemoglobin(a) and lyophilized seleno-methionine(b). The
Fe Kα and Kβ signals were confirmed by X-ray fluorescence of freeze-
5
Radiation Physics and Chemistry 181 (2021) 109334
dried hemoglobin powder, which is depicted in Fig. 6(a). The number
of photons per second is 0.73 at Kα and 0.15 at Kβ, and the ratio of Kβ/Kα
is 0.211. In the selenium Kα peak, the signal-to-noise ratio is 1.74, and
Kβ is 0.50. In the measurement of Fe X-ray fluorescence by the freeze-
dried hemoglobin, the position of the Kα, Kβ signal was correct, but
the signal-to-noise ratio was lower than the other samples because of the
high background noise. Seleno-methionine X-ray fluorescence spectrum
is depicted in Fig. 6(b). The number of photons per second is 5.70 at Kα
and 0.98 at Kβ, and the ratio of Kβ/Kα is 0.172. In the selenium Kα peak,
the signal-to-noise ratio is 7.95 and Kβ is 2.10. The reason for the lower
number of photons per second compared to the pure metal X-ray fluo­
rescence is that the content of metal in protein is only a few micrograms.
Experimental results show that tens of micrograms of metal material
can be measured accurately with various states of metal, such as solid
plate, solid powder, and lyophilized powder. The laser-induced X-ray
can replace the microfocus X-ray and the synchrotron source, which are
used as conventional micro X-ray fluorescence sources, and enables X-
ray fluorescence measurement with high resolution, high accuracy, and
high stability. Also, it can be applied in the field of material structure
analysis by using the short pulse width and high repetition rate of the
laser-induced X-ray. Based on the laser-induced X-ray light source, the
detector gating method measures high intensity for a short time in pi­
coseconds, so that enables radiographic imaging with improved signal to
noise ratio. With these advantages, the laser-induced X-ray source can be
used as a light source of X-ray fluorescence measurement with metal-
protein. By measuring the X-ray fluorescence of metal-protein mate­
rials, it is possible to trace the binding status between the bio-membrane
proteins and the metal-protein materials. By checking the reactivity
between the target metal-protein materials and the bio-membrane pro­
tein, this technique might be applied to the initial step for the screening
of drug candidates.
4. Conclusion
In this research, X-ray fluorescence of various metal compounds was
J. Kang and D.-K. Ko
Fig. 5. Laser-induced X-ray fluorescence spectrum of the selenium powder and the copper plate(a). Laser-induced X-ray fluorescence stability with selenium
powder(b)
Fig. 6. Laser-induced X-ray fluorescence spectrum of the lyophilized powder of human hemoglobin (a) and the laser-induced X-ray fluorescence spectrum of the
lyophilized powder of seleno-methionine(b)
measured with a laser-induced X-ray source. By using the wide brems­
strahlung energy region of the laser-induced X-ray source, it is possible
to analyze the X-ray fluorescence of various metal compounds that have
specific characteristic X-ray energy. To optimize the laser-induced X-ray
source intensity, we changed the pulse width, laser power, and pulse
focusing position. Al Kα and bremsstrahlung emission trends were
measured according to the change in laser intensity to find the optimum
conversion efficiency condition, and the difference of 10–20 μm spacing
between Al Kα maximum position and bremsstrahlung maximum posi­
tion was confirmed. We measured the X-ray fluorescence with copper
plate, selenium powder, lyophilized Hemoglobin, and lyophilized
seleno-methionine. As the content of metal atoms in the metal com­
pound decreases, the photon number per second of X-ray fluorescence
decreases, and the signal-to-noise ratio and accuracy of the Kα/Kβ ratio
also decrease. It was confirmed that the X-ray fluorescence measurement
was also possible with a few milligrams of metal compounds. The ac­
curate characteristic emission X-ray energy signal was measured with a
small amount of metal atoms. We think this technique has a potential to
be used as a step for high-throughput drug screening.
Author statements
Jungu Kang: Methodology, Investigation, Data Curation, Formal
analysis, Writing-Original Draft.
Do-Kyeong Ko: Supervision, Conceptualization, Methodology, Proj­
ect administration, Funding acquisition, Writing-Review & Editing.
6
Radiation Physics and Chemistry 181 (2021) 109334
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgments
This research was supported by the National Research Foundation of
Korea (NRF) (Grant Nos. NRF-2015R1A5A1009962 and NRF-
2016K1A3A1A61005950).
References
Alkhimova, M.A., Faenov, A. Ya, Pikuz, T.A., Skobelev, I. Yu, Pikuz, S.A., Nishiuchi, M.,
Sakaki, H., Pirozhkov, A.S., Sagisaka, S., Dover, N.P., Kondo, Ko, Ogura, K.,
Fukuda, Y., Kiriyama, H., Esirkepov, T., Bulanov, S.V., Andreev, A., Kando, M.,
Zhidkov, A., Nishitani, K., Miyahara, T., Watanabe, Y., Kodama, R., Kondo, K., 2018.
X-ray emission from stainless steel foils irradiated by femtosecond petawatt laser
pulses. IOP Conf. Series : J. Phys. Conf. 946 https://doi.org/10.1088/1742-6596/
946/1/012018, 012-018.
Azamoum, Y., Tcheremiskine, V., Clady, R., Ferré, A., Charmasson, L., Utéza, O.,
Sentis, M., 2018. Impact of the pulse contrast ratio on molybdenum Kα generation by
ultrahigh intensity femtosecond laser solid interaction. Sci. Rep. 8, 4119. https://
doi.org/10.1038/s41598-018-22487-3.
M. Barberio, S. Veltri, M. Sciscio, P. Antici, Laser-Accelerated proton beams as
diagnostics for cultural heritage, Sci. Rep. 7. 40415. https://doi.org/10.1038/srep
40415 .
J. Kang and D.-K. Ko
Cristoforetti, G., Anania, M.P., Faenov, A. Ya, Giulietti, A., Giulietti, D., Hansen, S.B.,
Koester, P., Labate, L., Levato, T., Pikuz, T.A., Gizzi, L.A., 2013. Spatially resolved
analysis of Kα x-ray emission from plasmas induced by a femtosecond weakly
relativistic laser pulse at various polarizations. Phys. Rev. 87 https://doi.org/
10.1103/physreve.87.023103, 023-103.
Feurer, T., Morak, A., Uschmann, I., Ziener, Ch, Schwoerer, H., Förster, E., Sauerbrey, R.,
2001. An incoherent sub-picosecond X-ray source for time-resolved X-ray-diffraction
experiments. Appl. Phys. B 72, 15–20. https://doi.org/10.1007/s003400000456.
Fourmaux, S., Kieffer, J.C., 2016. Laser-based Kα X-ray emission characterization using a
high contrast ratio and high-power laser system. Appl. Phys. B 122, 162. https://doi.
org/10.1007/s00340-016-6442-8.
Gibbon, P., Förster, E., 1996. Short-pulse laser–plasma interactions. Plasma Phys. Contr.
Fusion 38, 769–793. https://doi.org/10.1088/0741-3335/38/6/001.
Goryaev, A.A., Andreev, A.A., Platonov, K. Yu, Sedov, M.V., Zhavoronkov, N.I., 2014. X-
ray generation under oblique incidence of femtosecond laser pulses with kHz pulse
repetition rate on a metal foil with irregular relief. LASERS AND THEIR
APPLICATION 117 (6), 990–994. https://doi.org/10.1134/s0030400x14120091.
H€ofer, S., K€ampfer, T., F€orster, E., St€ohlker, T., Uschmann, I., 2016. Communication:
the formation of rarefaction waves in semiconductors after ultrashort excitation
probed by grazing incidence ultrafast time-resolved x-ray diffraction. Structural
dynamics 3. https://doi.org/10.1063/1.4963011, 051-101.
Iqbal, M., Urrehman, Z., Im, H., Son, J.G., Seo, O., Stiel, H., Nickles, P.V., Noh, D.Y.,
Janulewiez, K.A., 2014. Performance improvement of a Kerformance improvement
of a Kα source by high-resolution thin-layer-graphite spectrometer and a
polycapillary lens. Appl. Phys. B 116, 305–311. https://doi.org/10.1007/s00340-
013-5691-z.
Larsson, J., Allen, A., Bucksbaum, P.H., Falcone, R.W., Lindenberg, A., Naylor, G.,
Missalla, T., reis, D.A., Scheidt, K., Sjögren, A., Sondhauss, P., Wulff, M., Wark, J.S.,
2002. Picosecond X-ray diffraction studies of laser-excited acoustic phonons in InSb.
Appl. Phys. A 75, 467–478. https://doi.org/10.1007/s003390201421.
Li, Minghua, Huang, Kai, Chen, Liming, Yan, Wenchao, Tao, Mengze, Zhao, Jiarui,
Maa, Yong, Li, Yifei, Zhang, Jie, 2017a. Laser-driven powerful kHz hard x-ray
source. Radiat. Phys. Chem. 137, 78–82. https://doi.org/10.1016/j.
radphyschem.2016.01.042.
Radiation Physics and Chemistry 181 (2021) 109334
Li, Runze, Ashour, Omar A., Chen, Jie, Elsayed-Ali, H.E., Rentzepis, Peter M., 2017b.
Femtosecond laser induced structural dynamics and melting of Cu (111) single
crystal. An ultrafast time-resolved x-ray diffraction study. J. Appl. Phys. 121 https://
doi.org/10.1063/1.4975198, 055-102.
Morak, A., Kämpfer, T., Uschmann, I., Lübcke, A., Förster, E., Sauerbrey, R., 2006.
Acoustic phonons in InSb probed by time-resolved X-ray diffraction. Phys. Status
Solidi B 243 (No. 12), 2728–2744. https://doi.org/10.1002/pssb.200542387.
Nakano, H., Andreev, A.A., Limpouch, J., 2004. Femtosecond X-ray line emission from
multilayer targets irradiated by short laser pulses. Appl. Phys. B 79, 469–476.
https://doi.org/10.1007/s00340-004-1582-7.
Otendal, M., Tuohimaa, T., Hertz, H.M., 2007. Stability and debris in high-brightness
liquid metal-jet-anode microfocus x-ray sources. J. Appl. Phys. 101 https://doi.org/
10.1063/1.2423229, 026102.
Sokolowski-Tinten, Klaus, von der Linde, Dietrich, 2004. Ultrafast phase transitions and
lattice dynamics probed using laser-produced x-ray pulses. J. Phys. Condens. Matter
16. https://doi.org/10.1088/0953-8984/16/49/r04.
Tillman, C., Persson, A., Wahlstrim, C.-G., Svanberg, S., Herrlin, K., 1995. Imaging using
hard X-rays from a laser-produced plasma. Appl. Phys. B 61, 333–338. https://doi.
org/10.1007/bf01081532.
F.Valle Brozas, A. Crego, L. Roso, A. Peralta Conde, Laser-based X-ray and electron
source for X-ray fluorescence studies, Appl. Phys. B 122, 220. https://doi.org/10.100
7/s00340-016-6494-9 .
Yu, J., Jiang, Z., Kieffer, J.C., Krol, A., 1999. Hard x-ray emission in high intensity
femtosecond laser–target interaction. Phys. Plasmas 6, 13–18. https://doi.org/
10.1063/1.873372.
Zamponi, F., Ansari, Z., Korff Schmising, C.v., Rothhardt, P., Zhavoronkov, N.,
Woerner, M., Elsaesser, T., Bargheer, M., Trobitzsch-Ryll, T., Haschke, M., 2009.
Femtosecond hard X-ray plasma sources with a kilohertz repetition rate. Appl. Phys.
A 96, 51–58. https://doi.org/10.1007/s00339-009-5171-9.
Zhang, B.B., Sun, S.S., Sun, D.R., Tao, Y., 2014. Note: a novel normalization scheme for
laser-based plasma x-ray sources. Rev. Sci. Instrum. 85 https://doi.org/10.1063/
1.4896252, 096-110.