Gold nanoparticles as dehydrogenase mimicking nanozymes for estradiol degradation

Article abstract of DOI:10.1016/j.cclet.2019.05.062

Nanozyme catalysis has been mainly focused on a few chromogenic and fluorogenic substrates, while environmentally and biologically important compounds need to be tested to advance the field. In this work, we studied oxidation of estradiol (E2) in the presence of various nanomaterials including gold nanoparticles (AuNPs), nanoceria (CeO₂), Fe₃O₄, Fe₂O₃, MnO₂ and Mn₂O₃, and found that AuNPs had a dehydrogenase-mimicking activity to convert E2 to estrone (E1). This conversion was monitored using HPLC. The reaction was faster at higher pH and reached saturation at pH 8. Smaller AuNPs had a higher catalytic efficiency and 5 nm AuNPs were 4.8-fold faster than 13 nm at the same total surface area. Finally, we tried 17α-ethinylestradiol (EE2) as a substrate and found that 5 nm AuNPs can catalyze EE2 oxidation in the presence of H₂O₂. This work indicated that some nanomaterials can affect environmentally important hormones via oxidation reactions, and this study has expanded the scope of substrate of nanozymes.

Full text of DOI:10.1016/j.cclet.2019.05.062

G Model
CCLET 5030 No. of Pages 4
Chinese Chemical Letters xxx (2019) xxx–xxx
Contents lists available at ScienceDirect
Chinese Chemical Letters
journal homepage: www.elsevier.com/locate/cclet
Communication
Gold nanoparticles as dehydrogenase mimicking nanozymes
for estradiol degradation
Zijie Zhanga, Leslie M. Braggb, Mark R. Servosb, Juewen Liua,
*
a
Department of Chemistry, Waterloo Institute for Nanotechnology, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada
b
Department of Biology, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada
A R T I C L E I N F O
A B S T R A C T
Article history:
Nanozyme catalysis has been mainly focused on a few chromogenic and fluorogenic substrates, while
environmentally and biologically important compounds need to be tested to advance the field. In this
work, we studied oxidation of estradiol (E2) in the presence of various nanomaterials including gold
nanoparticles (AuNPs), nanoceria (CeO2), Fe3O4, Fe2O3, MnO2 and Mn2O3, and found that AuNPs had a
dehydrogenase-mimicking activity to convert E2 to estrone (E1). This conversion was monitored using
HPLC. The reaction was faster at higher pH and reached saturation at pH 8. Smaller AuNPs had a higher
catalytic efficiency and 5 nm AuNPs were 4.8-fold faster than 13 nm at the same total surface area. Finally,
we tried 17α-ethinylestradiol (EE2) as a substrate and found that 5 nm AuNPs can catalyze EE2 oxidation
in the presence of H2O2. This work indicated that some nanomaterials can affect environmentally
important hormones via oxidation reactions, and this study has expanded the scope of substrate of
nanozymes.
Received 25 April 2019
Received in revised form 30 May 2019
Accepted 30 May 2019
Available online xxx
Keywords:
Nanozymes
Estradiol
Gold nanoparticles
Catalysis
Environment
HPLC
© 2019 Chinese Chemical Society and Institute of Materia Medica, Chinese Academy of Medical Sciences.
Published by Elsevier B.V. All rights reserved.
Estrogens are the primary female sex hormones that play
critical roles in reproductive systems and secondary sex character-
istics [1,2]. For example, estradiol (E2), the major estrogenic
compound, is responsible for the regulation of estrous and
menstrual cycles [3]. However, accumulation of environmental
estrogens due to wastewater effluents and pharmaceuticals
discharges has raised prominent concerns [4,5]. For example,
estrogens from municipal wastewater were reported to feminiza-
tion of wild fish [6]. Exposure to low concentrations of
environmental estrogens interferes with human endocrine sys-
tems [7].
To minimize the hazard, many methods have been developed to
degrade estrogens through water treatment by using microorgan-
ism [8], ozone [9], ultrasound [10], UV and hydrogen peroxide or
their combinations [11]. Photocatalysis was also reported to be
capable of estrogen degradation [12]. For example, estradiol (E2)
was degraded using TiO2 combined with UV light [13]. Although
effective, these methods are complicated, high-cost, or need
energy input.
For example, gold nanoparticles (AuNPs) have glucose oxidase
activity [19,20], iron oxide nanoparticles have peroxidase activity
[21], and nanoceria has oxidase and other types of redox activities
[22,23]. With the advantages of high stability and low cost,
nanozymes are attractive in various applications ranging from
biosensor development [24–26] to nanomedicine [27–30]. So far,
most nanozymes have focused on some model chromogenic and
fluorogenic substrates [31–37], while their conversion of environ-
mentally and biologically important substrates is much less
explored [38–40]. For example, nanozymes have not yet been
tested for estrogen-based substrates. Part of the difficulty is
associated with the detection. Without an optical signal change,
more advanced techniques have to be used for the analysis.
In the present study, we screened six common nanozymes
using estrogen compounds as substrates to determine the
potential catalytic activity. We found that AuNPs can catalyze E2
oxidation to E1 in a simple aqueous solution without the need of
extra chemicals or energy input. Therefore, the AuNPs showed a
dehydrogenase-like activity in this case.
Many inorganic nanomaterials have enzyme-like activities
under ambient conditions, and they are called nanozymes [14–18].
Five common estrogen compounds were tested (Fig. 1A),
including three natural estrogens: estrone (E1), 17β-estradiol
(E2) and estriol (E3) (named as the number of hydroxyl group
contained), and two synthetic estrogens 17α-ethinylestradiol (EE2)
and bisphenol A (BPA). Since E2 is the most potent estrogen, we
first screened the nanomaterials with E2. AuNPs, nanoceria (CeO2),
* Corresponding author.
E-mail address: liujw@uwaterloo.ca (J. Liu).
https://doi.org/10.1016/j.cclet.2019.05.062
1001-8417/© 2019 Chinese Chemical Society and Institute of Materia Medica, Chinese Academy of Medical Sciences. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Z. Zhang, et al., Gold nanoparticles as dehydrogenase mimicking nanozymes for estradiol degradation, Chin.
Chem. Lett. (2019), https://doi.org/10.1016/j.cclet.2019.05.062

G Model
CCLET 5030 No. of Pages 4
2
Z. Zhang et al. / Chinese Chemical Letters xxx (2019) xxx–xxx
same time, no E1 was measured for the samples without AuNPs,
confirming the high catalytic activity of AuNPs (black trace,
Fig. 2C).
The 17β-OH group in E2 was converted to a keto in E1 (Fig. 2B),
and this is a typical dehydrogenation reaction (a type of oxidation).
This oxidation mechanism has been well studied in the AuNPs
catalyzed alcohol oxidation [44]. For example, AuNPs catalyze
primary alcohol oxidation to carboxylic acid and convert secondary
alcohol to be ketones [45]. AuNPs can also catalyze glucose
oxidation to gluconate [19,20]. In these reactions, AuNPs are
mimics of oxidases with ability of transporting electrons from
reducing substrates (e.g., glucose and E2) to the oxidizing agent,
oxygen. E2 also has a phenol hydroxyl group at the C3 position, and
this group was retained during the reaction (Fig. 2B). This fast
conversion of E2 to E1 may have significant environmental
implications. Since E1 has a far weaker activity as estrogen relative
to E2 [46], this study provides an effective method for E2
detoxification in aqueous by simply using AuNPs.
In the metabolic pathway, E2 oxidation to E1 is usually
catalyzed by 17β-hydroxysteroid dehydrogenases (17β-HSD)
[46]. Since 17β-HSD is an isoenzyme complex containing both
reductases and dehydrogenases, the reaction is reversible [47], i.e.,
oxidation from E2 to E1 and reduction from E1 to E2. To test
whether the AuNPs have such reversible activity, we then mixed E1
with the AuNPs and analyzed the products in the same method
(Fig. 2D). From the LC-MS spectra, aside from the added E1 with
peaks at ꢀ8.3 min, no E2 or other compounds were identified (the
small peaks were from the standards). Therefore, the AuNPs
worked only as an oxidase (dehydrogenase) mimic for E2
degradation and the reaction is irreversible (Fig. 2B).
To further optimize and understand the reaction, we then
mixed the AuNPs with E2 in different pH conditions and the results
are summarized in the Fig. 3A. The yield of E2 oxidation product
was higher at high pH. The maximum production of E1 was
obtained at pH 8 reaching ꢀ2.7 mg/L. As controls, basically no E1
was measured for all of the samples without AuNPs. We also tested
the reaction in phosphate buffers of different pH and a higher
activity was still observed at higher pH (Fig. 3B). If tested in
different buffers but all at pH 7, the activities were similar (Fig. 3C).
Therefore, the oxidation was mainly affected by pH and the type of
buffer was not important.
Fig. 1. (A) Chemical structures of the estrogens used in this study. BPA was also
included as an environmental contaminant for analysis. (B) The LC-MS spectra of E2
only sample (Ctrl) and E2 mixed with different nanozymes for 20 min in HEPES
buffer (20 mmol/L, pH 7) at room temperature. The other peaks were from the
spiked standards to assign the peaks.
Fe3O4, Fe2O3, MnO2 and Mn2O3 nanoparticles were chosen for
screening. These nanomaterials were spherical and their sizes
ranged from ꢀ5 nm to 50 nm, with nanoceria being the smallest of
ꢀ5 nm and Mn2O3 being the largest of ꢀ50 nm [41]. These
nanoparticles were respectively incubated with E2. Then after
centrifugation to precipitate the nanoparticles, the supernatants
were quantitatively measured by using LC-MS. To determine the
retention time of each estrogen in the LC-MS spectra, a set of
standards containing all these estrogens were also added to each
sample before analysis, and these peaks can be observed from the
control sample without any nanomaterial added (red trace,
Fig. 1B). For the nanomaterial treated samples, only the AuNP
added sample produced a new peak (at ꢀ8.3 min retention time),
suggesting that it could convert E2 to another compound.
Since only the AuNPs showed activity, we focused our study on
it. The AuNPs used here were ꢀ13 nm in diameter (Fig. 2A). From
the retention time of each estrogen on the control spectra by
running the standard samples (Fig. 1B) [42,43], E2 had a peak at
ꢀ7.8 min and the converted product at ꢀ8.3 min was identified to
be E1. The production of E1 was also confirmed from the mass
spectrometry (MS) data (Fig. S1 in Supporting information). Other
estrogens with small peaks were only from the standards
(Fig. 1B). Therefore, the AuNPs catalyzed the conversion of E2
to E1 (Fig. 2B).
To determine the reaction rate, we then measured the kinetics
by plotting the converted E1 as a function a reaction time (red
trace, Fig. 2C). The reaction was very fast, since the converted E1
reached to the maximum of ꢀ2.0 mg/L within just 10 min, and the
first-order rate constant was calculated to be 0.22 min-1. At the
We reason that the reaction of E2 oxidation released protons so
that producing E1 was favored at alkaline conditions. This
Fig. 2. (A) A TEM photo of 13 nm AuNPs. (B) AuNPs catalyze oxidation of E2 to E1
and the reaction is irreversible. The red circled group of E2 is 17β-OH. (C) Kinetics of
the converted E1 from E2 oxidation with or without AuNPs. (D) The LC-MS spectra
of E1 with or without mixed with AuNPs for 1 h. The small peaks of E2 and E3 were
from the spiked standards to assign the peaks. E2 or E1 (0.1 mmol/L) and 13 nm
AuNPs (5 nmol/L) were used for the reactions in HEPES buffer (20 mmol/L, pH 7) at
room temperature.
Fig. 3. Concentrations of obtained E1 from E2 oxidations catalyzed by the AuNPs in
(A) different buffers with different pH, (B) phosphate buffers with different pH, (C)
different buffers with same pH at 7. (D) The recycle tests of E2 oxidations catalyzed
by AuNPs. E2 (0.1 mmol/L) and 13 nm AuNPs (5 nmol/L) were used for the reactions
at room temperature.
Please cite this article in press as: Z. Zhang, et al., Gold nanoparticles as dehydrogenase mimicking nanozymes for estradiol degradation, Chin.
Chem. Lett. (2019), https://doi.org/10.1016/j.cclet.2019.05.062

G Model
CCLET 5030 No. of Pages 4
Z. Zhang et al. / Chinese Chemical Letters xxx (2019) xxx–xxx
3
phenomenon was also observed by using manganese oxide to
catalyze E2 oxidation in soil [48]. In human metabolic pathway,
dehydrogenase catalyzing E2 oxidation to E1 also produces protons
that are taken by the enzyme cofactor NADP [49,50]. An interesting
comparison can be made is glucose oxidation catalyzed by AuNPs,
where the AuNPs are mimics of glucose oxidase [20]. In this
reaction, the oxidation rate is also higher at higher pH. Protons
were produced during the reaction and the decrease of pH
inhibited the activity [51]. Electrostatic force may be favorable for
the catalysis, since E2 has a pKa of ꢀ10.7 and it is positive charged
in pH range [52], while the AuNPs are negative charged [53,54].
Previous mechanism studies suggested that more active species
were on gold surfaces compared with other metal nanomaterials.
The reaction might follow the Eley–Rideal mechanism that the
adsorbed E2 on the surface reacts with oxygen in the solution
[20,55].
The concentration of the AuNPs was 5 nmol/L. With a yield of
2.7 mg/L of E1 (10 mmol/L), each AuNP converted 2000 E2 to E1
during this reaction time of 20 min. Therefore, this is a quite
efficient multi-turnover nanozyme. The AuNPs can be reused for
multiple catalysis cycles by simple centrifugation and re-disper-
sion (Fig. 3D).
In general, smaller nanoparticles are more active due to
larger specific surface area. In addition, smaller particles are
more prone to form defect sites and these sites are often more
catalytically active. We then used smaller AuNPs of 5 nm (Fig. S2
in Supporting information for TEM) to catalyze E2 oxidation and
compared it with the 13 nm AuNPs. With same molar
concentration (5 nmol/L), 5 nm AuNPs converted ꢀ1.4 mg/mL
of E1, which was slightly less than that of 13 nm AuNPs of ꢀ2 mg/
mL (Fig. 4A). Since the surface area of 5 nm AuNP is 6.7-fold
smaller than that of a 13 nm AuNP, when comparing them based
on same surface area, the oxidation yield of 5 nm AuNPs was
ꢀ4.8-fold higher than that of 13 nm AuNPs (Fig. 4B). Therefore,
5 nm AuNP is a more effective catalyst for E2 degradation likely
due to it has a higher active site density.
Fig. 5. LC-MS spectra of EE2 oxidation catalyzed by 5 nm and 13 nm AuNPs (A)
without, and (B) with 10 mmol/L H2O2 added in HEPES buffer (20 mmol/L, pH 7) at
room temperature for 1 h.
In conclusion, with more and more nanomaterials being
produced for various applications, their environmental impact
will have to be carefully studied. In this work, we screened six
commonly used inorganic nanomaterials for reaction with the E2
estrogens. AuNPs were found capable of catalyzing E2 oxidation by
simply mixing in aqueous solutions, while the tested metal oxide
nanoparticles had no activity. The reaction has higher oxidation
rate in alkaline conditions since the produced proton during
oxidation. The AuNPs are mimics of dehydrogenases in human
metabolic pathway for E2 degradation. The smaller size of AuNPs (
e.g. 5 nm) has more effective catalytic activity due to its larger
surface area for substrates accessibility. This work has expanded
the scope of the nanozyme field by testing a new set of
environmental substrates. The reaction may also be extended to
some signal transduction mechanisms for biosensor development
[56,57]. We expect more developments in this front to show
practical environmental and biological applications of nanozymes
in the near future.
Acknowledgment
Funding for this work was from the Natural Sciences and
Engineering Research Council of Canada (NSERC).
After confirming E2 oxidation catalyzed by the AuNPs, we then
wanted to try EE2. We chose EE2 since it is a synthetic derivative
of E2 and wildly used as estrogen medication. EE2 was mixed
with the AuNPs for reaction and the products were analyzed by
LC-MS in the same method (Fig. 5A). However, no products were
observed indicating that EE2 did not degrade. Compared to E2, the
carbon has an additional 17α-ethynyl group (Fig. 1A), making it a
tertiary alcohol and more difficult to oxidize [46]. Next, we added
H2O2 to the reaction. From the spectra (Fig. 5B), we found only a
trace amount of E1 was produced with by 5 nm AuNPs while not
for the 13 nm AuNPs. No E1 was found when only added H2O2
without AuNPs (Fig. S3 in Supporting information). In this case, the
AuNPs were mimics of peroxidase activity in the reaction. Besides
EE2, we also tried E3 and BPA as substrates for oxidation. However,
no oxidation products were detected (Fig. S4 in Supporting
information).
Appendix A. Supplementary data
Supplementarymaterialrelatedtothisarticlecanbefound, inthe
online version, at doi:https://doi.org/10.1016/j.cclet.2019.05.062.
References
[1] C.J. Gruber, W. Tschugguel, C. Schneeberger, et al., N. Engl. J. Med. 346 (2002)
340–352.
[2] H.G. Burger, Fertil. Steril. 77 (2002) 3–5.
[3] V.N. Luine, Horm. Behav. 66 (2014) 602–618.
[4] M. Adeel, X. Song, Y. Wang, et al., Environ. Int. 99 (2017) 107–119.
[5] T.F.T. Omar, A. Ahmad, A.Z. Aris, et al., TrAC-Trends Anal. Chem. 85 (2016) 241–259.
[6] R.N. Tanna, G.R. Tetreault, C.J. Bennett, et al., Environ. Toxicol. Chem. 32 (2013)
1981–1991.
[7] M. Giulivo, M.L. de Alda, E. Capri, et al., Environ. Res. 151 (2016) 251–264.
[8] C. Zhang, Y. Li, C. Wang, et al., Crit. Rev. Environ. Sci. Technol. 46 (2016) 1–59.
[9] F. Itzel, K.S. Jewell, J. Leonhardt, et al., Sci. Total Environ. 624 (2018) 1443–1454.
[10] A. Roudbari, M. Rezakazemi, AMB Express 8 (2018) 91.
[11] Z. Zhang, Y. Feng, Y. Liu, et al., J. Hazard. Mater. 181 (2010) 1127–1133.
[12] K. Sornalingam, A. McDonagh, J.L. Zhou, Sci. Total Environ. 550 (2016) 209–224.
[13] M.V. de Liz, R.M. de Lima, B. do Amaral, et al., J. Braz. Chem. Soc. 29 (2018) 380–389.
[14] Y. Huang, J. Ren, X. Qu, Chem. Rev. 119 (2019) 4357–4412.
[15] Y. Zhou, B. Liu, R. Yang, et al., Bioconj. Chem. 28 (2017) 2903–2909.
[16] J. Wu, X. Wang, Q. Wang, et al., Chem. Soc. Rev. 48 (2019) 1004–1076.
[17] S. Lin, H. Wei, Sci. China-Life Sci. 62 (2019) 710–712.
[18] G. Zhang, R. Wang, G. Li, Chin. Chem. Lett. 29 (2018) 687–693.
[19] W. Luo, C. Zhu, S. Su, et al., ACS Nano 4 (2010) 7451–7458.
[20] M. Comotti, C. Della Pina, R. Matarrese, et al., Angew. Chem. Int. Ed. 43 (2004)
5812–5815.
[21] L. Gao, J. Zhuang, L. Nie, et al., Nat. Nanotechnol. 2 (2007) 577–583.
[22] A. Asati, S. Santra, C. Kaittanis, et al., Angew. Chem. Int. Ed. 48 (2009) 2308–2312.
[23] M. Liu, Z. Li, Y. Li, J. Chen, Q. Yuan, Chin. Chem. Lett. 30 (2019) 1009–1012.
[24] B. Liu, J. Liu, Nano Res. 10 (2017) 1125–1148.
Fig. 4. (A) Concentration of converted E1 from E2 oxidation catalyzed by 5 nmol/L
of 13 nm or 5 nm AuNPs. (B) The normalized activity based on the same surface area.
For this experiment, 0.1 mmol/L E2 and 5 nmol/L AuNPs were used in HEPES buffer
(20 mmol/L, pH 7) at room temperature.
[25] Y. Hu, H. Cheng, X. Zhao, et al., ACS Nano 11 (2017) 5558–5566.
Please cite this article in press as: Z. Zhang, et al., Gold nanoparticles as dehydrogenase mimicking nanozymes for estradiol degradation, Chin.
Chem. Lett. (2019), https://doi.org/10.1016/j.cclet.2019.05.062

G Model
CCLET 5030 No. of Pages 4
4
Z. Zhang et al. / Chinese Chemical Letters xxx (2019) xxx–xxx
[26] T. Yu, B. Liu, J. Liu, J. Anal. Test. 1 (2017) 2.
[43] M.J. Arlos, R. Liang, M.M. Hatat-Fraile, et al., J. Hazard. Mater. 318 (2016) 541–550.
[44] C. Della Pina, E. Falletta, L. Prati, et al., Chem. Soc. Rev. 37 (2008) 2077–2095.
[45] T. Mallat, A. Baiker, Annu. Rev. Chem. Biomol. Eng. 3 (2012) 11–28.
[46] H. Kuhl, Climacteric 8 (2005) 3–63.
[27] L.Z. Gao, K.L. Fan, X.Y. Yan, Theranostics 7 (2017) 3207–3227.
[28] X. Wang, Y. Hu, H. Wei, Inorg. Chem. Front. 3 (2016) 41–60.
[29] Z. Liu, X. Qu, Sci. China-Life Sci. 62 (2019) 150–152.
[30] L. Yu, Y. Chen, H. Chen, Chin. Chem. Lett. 28 (2017) 1841–1850.
[31] Z. Zhang, X. Zhang, B. Liu, et al., J. Am. Chem. Soc. 139 (2017) 5412–5419.
[32] B. Liu, Z. Huang, J. Liu, Nanoscale 8 (2016) 13562–13567.
[33] Y. Biniuri, B. Albada, M. Wolff, et al., ACS Catal. 8 (2018) 1802–1809.
[34] M. Vazquez-Gonzalez, R.M. Torrente-Rodriguez, A. Kozell, et al., Nano Lett. 17
(2017) 4958–4963.
[47] M.M. Miettinen, M.V. Mustonen, M.H. Poutanen, et al., Biochem. J. 314 (1996)
839–845.
[48] G.D. Sheng, C. Xu, L. Xu, et al., Environ. Pollut. 157 (2009) 2710–2715.
[49] L.L. Engel, E.V. Groman, Human placental 17β-estradiol dehydrogenase:
Characterization and structural studies, Proceedings of the 1973 Laurentian
Hormone Conference, Elsevier, 1974 pp. 139–169.
[35] M.S. Hizir, M. Top, M. Balcioglu, et al., Anal. Chem. 88 (2016) 600–605.
[36] D. Li, B. Liu, P.J.J. Huang, et al., Chem. Commun. 54 (2018) 12519–12522.
[37] X.L. Ren, M.Q. Wang, X. He, et al., Chin. Chem. Lett. 29 (2018) 1865–1868.
[38] H. Liang, F. Lin, Z. Zhang, et al., ACS Appl. Mater. Interfaces 9 (2017) 1352–1360.
[39] C.I. Wang, W.T. Chen, H.T. Chang, Anal. Chem. 84 (2012) 9706–9712.
[40] J.H. Xu, C.G. Hu, S.S. Hu, Chin. Chem. Lett. 20 (2009) 1248–1250.
[41] B. Liu, J. Liu, ACS Appl. Mater. Interfaces 7 (2015) 24833–24838.
[42] M.J. Arlos, W.J. Parker, J.R. Bicudo, et al., Sci. Total Environ. 610 (2018) 1103–1112.
[50] R. Breton, D. Housset, C. Mazza, et al., Structure 4 (1996) 905–915.
[51] N.J. Lang, B. Liu, J. Liu, J. Colloid Interface Sci. 428 (2014) 78–83.
[52] K. Lewis, R. Archer, Steroids 34 (1979) 485–499.
[53] A. Kumar, S. Mandal, P. Selvakannan, et al., Langmuir 19 (2003) 6277–6282.
[54] J. Liu, Phys. Chem. Chem. Phys. 14 (2012) 10485–10496.
[55] Y. Lin, J. Ren, X. Qu, Adv. Mater. 26 (2014) 4200–4217.
[56] Z. Qing, L. Zhu, S. Yang, et al., Biosens. Bioelectron. 78 (2016) 471–476.
[57] D. Wang, B. Huang, J. Liu, et al., Biosens. Bioelectron. 102 (2018) 389–395.
Please cite this article in press as: Z. Zhang, et al., Gold nanoparticles as dehydrogenase mimicking nanozymes for estradiol degradation, Chin.
Chem. Lett. (2019), https://doi.org/10.1016/j.cclet.2019.05.062