P.
Kumar
a,
M.
Parashar
a,
K.
Chauhan
a,
N.
Chakraborty
b,
S.
Sarkar
b,
A.
Chandra
b,
N. S.
Das
c,
K. K.
Chattopadhyay
b,
A.
Ghoari
d,
A.
Adalder
d,
U. K.
Ghorai
d,
S.
Saini
e,
D.
Agarwal
e,
S.
Ghosh
e,
P.
Srivastava
e and
D.
Banerjee
*a
aThin Film and Nanotechnology Laboratory, Faculty of Engineering and Computing Sciences, Teerthanker Mahaveer University, Moradabad, UP 244001, India. E-mail: nilju82@gmail.com
bThin Film and Nanoscience Laboratory, Department of Physics, Jadavpur University, Kolkata, West Bengal 700032, India
cDepartment of Basic Science and Humanities, Techno International Batanagar, Maheshtala, Kolkata 700141, India
dDepartment of Industrial Chemistry, Ramakrishna Mission Vidyamandira, Belur Math, Howrah-711202, India
eDepartment of Physics, Indian Institute of Technology, Hauz Khas South West, Delhi, 110016, India
First published on 30th October 2023
The current article presents a huge enhancement in the field emission characteristics of zinc oxide (ZnO) micro/nanorods by nickel doping. The synthesis of pure and nickel-doped zinc oxide (ZnO) micro/nanorods was done by a simple low-temperature chemical method. Both the as-prepared pure and doped samples were analyzed by X-ray diffraction and electron microscopy to confirm the proper phase formation and the developed microstructure. UV-vis transmittance spectra helped in determining the band gap of the samples. Fourier-Transform Infrared Spectroscopy (FTIR) spectra showed the different bonds present in the sample, whereas X-ray Photoelectron Spectroscopy (XPS) confirmed the presence of nickel in the doped sample. Photoluminescence (PL) spectra showed that after doping, the band-to-band transition was affected, whereas defect-induced transition had increased significantly. After the nickel doping, contact angle measurement revealed a significant decrease in the sample's surface energy, leading to a remarkably high water contact angle (within the superhydrophobic region). Simulation through ANSYS suggested that the doped sample has the potential to function as an efficient cold emitter, which was also verified experimentally. The cold emission characteristics of the doped sample showed a significant improvement, with the turn-on field (corresponding to J = 1 μA cm−2) reduced from 5.34 to 2.84 V μm−1. The enhancement factor for the doped sample reached 3426, approximately 1.5 times higher compared to pure ZnO. Efforts have been made to explain the results, given the favorable band bending as well as the increased number of effective emission sites.
As mentioned earlier, ZnO has the potential to function as a good cold emitter. Thus, this is why there are considerable reports related to the cold emission properties of ZnO. For instance, Q. H. Li et al. reported that tetrapod-like nanostructures are promising candidates for stable FE with a fluctuation of less than 2% even after 72 hours.23 The maximum current density achieved in their work was 18 mA cm−2. A study on the thermo-enhanced FE phenomenon from ZnO nanowires was reported by Zhang et al. They reported that under a constant applied electric field, the field emission current exhibited a nearly hundred-fold increase accompanied by a significant decrease in the turn on field (ETO) as the temperature increased from 323 K to 723 K.24 Y. Lv et al. observed that brush-like ZnO nanowire arrays deposited over carbon cloth offered a remarkably low ETO of 1.02 V μm−1.25 The system also offered an exceptionally high field enhancement factor (β) of 64438. In the previous year, J. Singh and A. Srivastava synthesized ZnO nanostructures using the pulsed laser deposition technique. They observed the stable FE characteristics of the aligned ZnO with a low threshold field (ETh) of 2.40 V μm−1.26
In the era of nano-science and technology, it is a very common means that the performance of a specific material in a particular field of application is made significantly better by suitable doping, by functionalization or by making suitable hybrids or composites.
Similar instances can also be found in the cold emission application of ZnO. In this regard, the work that addressed the etching of ZnO nanorods by Ar plasma, as reported by Ji et al., is worth mentioning.27 Apart from the Ar treatment, ZnO was also coated with a layer of AlN to achieve favorable band bending. Here, ETO exhibited a remarkable 42% reduction to 6.8 V μm−1, and the hybrid sample reached the highest current density value of 4.1 mA cm−2.
Another report that addressed the FE from graphene-wrapped ZnO came from X. Wang et al. It was shown that the wrapping of the wrinkled graphene structure enhances the FE characteristics of ZnO significantly,28 resulting in ETO and ETh values of 1.63 and 3.12 V μm−1, respectively.
J. Chen et al. employed a similar approach of aluminium doping through electron beam evaporation to enhance the field emission characteristics of ZnO.29 The system gave the best emission characteristics with ETO as low as 6.21 V μm−1 for optimized layer thickness.
Young et al. investigated the field emission properties of hydrothermally synthesized Ag-doped ZnO.30 When the sample was irradiated with UV light, the doped ZnO gave the lowest ETO with a value of 2 V μm−1 associated with a current density of 1 μA cm−2.
An ETO value of 4.7 V μm−1 with a β value of 1686 was obtained from a ZnO rod deposited over silicon pre-patterned by Au.31 In the same study, Lu et al. showed that the screening effect could significantly deteriorate the emission performance of the system. A. Nagar et al. employed the sol–gel method for the synthesis of a ZnO nanostructure.32 It was shown that the geometry of the flower-like structure significantly affects the emission characteristics and related parameters such as current density, ETO, and β.
The synthesis of Al- and Cu-doped ZnO by a hydrothermal method with varying dopant concentrations was reported by Liu et al.33 Both systems offered good FE characteristics. However, the best result with a turn-on field of 1.51, 1.80 V μm−1 from 6% Al- and 4% Cu-doped samples, respectively, was obtained only after wrapping the system with a graphene sheet. S. J. Young and Y. L. Chu developed a system of Pd nanoparticle decorated ZnO quantum dots and quantum wires.34 In their case, including Pd did not have a marked effect in changing the turn-on field, but the decorated system showed more than a 100% increase in the enhancement factor. Jun Chen obtained a very high emission current of 20 mA et al. from their developed system of indium-doped ZnO.35 The high emission current was attributed to the high crystallinity and conductivity of the doped sample.
However, the existing studies, though they sometimes provide good FE results, lack other important features required for proper industry-scale applications.
For instance, Li et al.23 used a high-temperature synthesis process, while Zhang et al.24 needed additional thermal energy to give the best emission result, indicating that the process cannot be considered to be pure tunneling. The method reported by Ye et al. also involves a high-temperature process, and the value of ETO was also quite high.27 The work led by Cao demonstrated that the screening effect could reduce the FE characteristics of ZnO.31 Though this has fundamental importance, it lacks practical application. The work in 34 showed that including Pd did not affect the emission characteristics much. Though doped ZnO gives a very interesting and useful FE result, it has the same drawback with a high thermal energy required in the synthesis process.35
It is also rather surprising that even after the prediction of ANSYS-based simulation regarding the good potential of nickel-doped ZnO as a field emitter, there are a negligible number of reports regarding the same. Only Chang et al. developed Ni-doped ZnO by a thermal evaporation process, which is a high-temperature synthesis approach. For the optimized system, the required temperature was as high as 1100 °C.36 Their as-grown different nanostructure was able to give moderately good FE characteristics and offered a turn-on field of around 1.53 V μm−1, which is rather low compared to the present work. But it is also to be noted that they have defined the turn-on field at a current density value of 0.1 μA cm−2, which is 10 times inferior compared to the present work.
In 2014, Rana et al. reported on the cold emission of Ni-doped ZnO prepared by the wet chemical method. However, they achieved only an 8% reduction of the turn-on field in their 5% Ni-doped ZnO.37 Also, the article remained almost silent regarding the interaction between the sample and water, where excess hydrophilicity of any electronic components is a considerable problem.
Keeping this in mind, the present work deals with the synthesis of ZnO nanorods (ZnO NR) by a KMnO4 seeding-mediated wet chemical route and doping it with transition metal elements like nickel. The as-prepared pure and hybrid samples were characterized by various sophisticated techniques to confirm their phase, microstructure, doping, and bonding information. It is shown that both the samples are well luminescent with emission wavelengths lying in the blue/cyan range. It is also seen that after doping, the characteristics of the sample readily switch over from hydrophilic to superhydrophobic, and the cold emission characteristics are drastically enhanced, with ET and β undergoing a huge enhancement of 2 and 48 times, respectively. As far as the authors are concerned, this is the first report of a superhydrophobic Ni-doped ZnO-based cold emitter. The enhancement has been explained in terms of field enhancement on the favorable nanostructure, favorable band bending, and an increased number of defects that come from oxygen vacancies created after Ni doping.
The current article is expected to be a valuable addition to the related literature.
The pure and doped samples were named sample A and B, respectively.
The as-prepared pure and doped samples were analyzed by using an X-ray diffractometer (XRD, BRUKER D8 Advance), field emission scanning electron microscope (FESEM, JEOL 6340F FEG-SEM), UV-vis spectrophotometer (Shimadzu UV-3600), Fourier transform infrared spectrometer (Shimadzu IR Affinity-1S FTIR) and photoluminescence spectrophotometer (PL, Shimadzu RF-5301 Spectrofluorophotometer). The surface energy of the samples was calculated by measuring the contact angle of two liquids of known polarity (OCA 15EC, Data Physics). The field emission characteristics of the samples were tested by using a laboratory-assembled field emission setup.
The average crystallite size (D) of the synthesized sample can be calculated using Scherrer's formula as given below:
(1) |
Fig. 2 XPS (a) survey scan of (b) O 1s and (c) Zn 2p spectra of both sample A and B; (d) Ni 2p high resolution scan of sample B. |
The presence of Ni can clearly be seen in Fig. 2d, where two main peaks centered at 855.29 and 873.46 eV are the signatures of Ni 2p3/2 and 2p1/2 levels. These two peaks are accompanied by two satellite peaks at 861.25 and 879.82 eV.40 These satellite peaks arise due to the difference in the electronic configuration between the ground state and excited state during photoelectron emission. The shakeup lines indicate a charge-transfer transition from O (2p) to Ni (3d), resulting in hybridization.41 When quantification analysis is done, it is revealed that the Ni is present with an atomic percentage of 1.76 and a weight percentage of 4.28. All these together confirm the successful doping of ZnO nanostructures.
From Fig. 3(a–c), it is seen that one-dimensional structures with a diameter of around 400 nm and length of the order of several micro-meters have been formed uniformly. The hexagonal structures of the individual rod can clearly be seen, with the width of each side lying between 100 and 120 nm.
When ZnO gets doped with Ni, it is seen that (Fig. 3(d–f)) though the structure of the sample was still one dimensional, it undergoes distortion, and 2 D flake-like structures with sharp edges begin to form. These structures are expected to contribute to further field enhancement and better field emission characteristics, which will be discussed in the coming section.
Fig. 4 Transmittance spectra (a) and Tauc plot (b) of both the pure and doped samples (inset (b): the Tauc plot of sample B). |
The standard Tauc formula is used to calculate the absorption coefficient (α).
(2) |
(αhv)2 = A(hv − Eg) | (3) |
It is clearly seen that both samples have few peaks common with slight shifting. For instance, the peaks at 436, 506, and 1276 cm−1 shift to 441, 508, and 1275 cm−1 in the case of sample B. Sample A has other prominent peaks at 682 and 935 cm−1 and a band between 1436 and 1762 cm−1. In the case of sample B, the information can be seen at 666, 811, 1088, 1529, 1685, and 1757 cm−1. The band in sample A forms due to the overlapping of different C–H, C–C, and CO bonds, either absorbed or formed during synthesis on the ZnO surface.42 Of all these, peaks at 436 (or 441) and 506 (or 508) cm−1 are the signatures of the Zn–O stretching bond,43 of which the peak corresponding to 506 cm−1 is Raman active.44 A very strong peak at 935 cm−1 in the case of sample A signifies a considerable number of O–H groups in sample A.45 Thus there is a probability that this OH group could react with external water droplets, forming a hydrogen bond resulting in its hydrophilic nature, which will be discussed in the coming section. The peak at 682 cm−1 in sample A also confirms the presence of a Zn–O bond, which may shift and overlap with the peak at 666 cm−1, as seen in sample B.46 This peak at 666 cm−1, seen in the case of sample B may probably be due to the Ni–O stretching vibrational mode.47 This indirectly confirms the presence of Ni in sample B. The transmittance peak at 1275 cm−1 shown in both samples probably comes from the carbonaceous impurities48 or the Si substrate.49 The peaks at 811, 1685, 1088, 1529, and 1757 cm−1 appear in the spectra of sample B and bear the signatures of C–H out of plane bending, CO stretching C–C/CC stretching, and CO stretching modes.46,50–52
Regarding the second peak sample A has an intensity not as high as that of sample B, and in the case of sample B, the peak with higher intensity has much higher broadening. Thus the peak basically gets converted to a broad band. This confirms that sample B is much more defect rich, and it would not be wrong to assume that the defect is mainly due to the oxygen vacancy, as confirmed by the XPS study. This has a marked effect on the cold emission characteristics, which will be discussed in the coming section. The CIE chromaticity diagrams for the doped and un-doped samples are shown in Fig. 7a and b. The CIE (x, y) coordinate of sample A is (0.26312, 0.27115), and for B, it is (0.28452, 0.29641). From the chromatography diagram, it is clearly seen that after doping, the emission color moves toward cyan from the blue region.
Fig. 8 Response of the sample A towards (a) water, (c) glycerol and sample B towards (b) water and (d) glycerol. |
Also, the first principles study done by Wang et al.53 in explaining the hydrophobic properties of aluminium doped ZnO can be mentioned here to understand the change in the hydrophobic nature of the sample. Experimentally the group has shown that when pure ZnO is doped with Al the hydrophobicity of the doped sample increased significantly and they claimed that the preferential growth of ZnO crystals can be attributed to the change in the hydrophobic nature of the sample. They have shown that after doping, the intensities of both the XRD peaks associated with (100) and (002) planes increased significantly as in the present case of nickel doped ZnO.
It is reported that in the case of the (002) plane of ZnO, the oxygen atom residing on the surface forms a bond with the hydrogen atom of water with a bond length O(ZnO)–H(H2O) of 1.893 A, and a O(ZnO)–H(H2O)–O(H2O) angle of 172.2°. In contrast, on ZnO (100) surfaces, oxygen of water forms a strong bond with the Zn atom, to gain the minimum energy configuration and the corresponding Zn(ZnO)–O(H2O) distance is 2.094 A with a Zn(ZnO)–O(H2O)–H(H2O, up) angle of 118.0°. The higher bond length and lower bond angle always play an important role in gaining enhanced hydrophobicity. Thus, the relative abundance of these two crystal planes may be the key reasons for such enhanced hydrophobicity (in the present case it is ZnO(100)/ZnO(002) = 1.27 and 1.42 respectively for pure and doped samples) along with the other parameters like changes in microstructures, roughness or porosity. However, these needs further studies to be verified.
These contact angle data can be used to calculate both the polar (Sp) and dispersive (Sd) components of the surface energy using the equation below:54
(4) |
Here, we have taken Splv = 51.0 mJ m−2 and Sdlv = 21.8 mJ m−2 for water and Splv = 30.0 mJ m−2 and Sdlv = 34.0 mJ m−2 for glycerol to calculate the surface energy; all the results have been summarized in Table 1.55 The result shows that the polar part of the surface energy governs the water contact, as reported previously.
Sample | Contact angle (°) | Surface energy mN m−1 | |||
---|---|---|---|---|---|
Water | Glycerol | Polar | Dispersive | Total | |
A | 57.65 | 41.29 | 15.94 | 34.36 | 50.29 |
B | 150 | 138.81 | 0.28 | 3.45 | 3.73 |
The primary factor contributing to the reduced water contact angle on surface A, as indicated by the FTIR spectra, could be attributed to the presence of a significant number of OH groups. These groups interact with water by forming unwanted hydrogen bonds resulting in hydrophilicity in this sample.
Fig. 9e–h show the same parameters for sample B. Fig. 9c, d, g and h are basically the same duo of field line distribution where transparency is introduced only to show the concentration of the field over the sharp tips/edges of the nanostructures.
It is clearly seen that for the doped sample, the intensity of the applied electric field is much higher, and at the same time, the field distribution is much more favorable for the doped sample to function as a cold emitter.
Propelled by the positive theoretical result, FE measurements were carried out using a laboratory-assembled field emission setup. The measurements were carried out using a diode configuration consisting of a cathode (the film under test) and a stainless steel anode mounted on a liquid nitrogen-trapped rotary-diffusion high vacuum chamber with an appropriate chamber baking arrangement. The measurements were performed at a base pressure of ∼10−7 mbar. The inter-electrode distance was kept as high as 300 μm. The macroscopic applied electric field can be obtained by dividing the applied voltage by the inter-electrode distance. Theoretically, the emission current and the macroscopic electric field are related to each other by the well-known Fowler–Nordheim equation56
I = AatF−2ϕ−1(βE)2exp{−bvFϕ3/2/βE} | (5) |
ln{J/E2} = ln{tF−2aϕ−1β2} − [{vFbϕ3/2β−1}/E] | (6) |
Hence, the plot of ln {J/E2} vs. 1/E should be a straight line, and its slope and intercept give valuable information about the enhancement factor, local work function, etc. An experimental F–N plot has been modelled, which can be expressed as
ln{J/E2} = ln{raϕ−1β2} − [{sbϕ3/2β−1}/E] | (7) |
It is seen that the emission current density has a huge enhancement in the case of the doped sample. Also, the turn-on field (ETO), which is defined as the field required to produce a current density of 1 μA cm−2, reduced from 5.34 to 2.84 V μm−1. The value 2.84 V μm−1 may sometimes seem comparable to the reported value, but it is to be kept in mind that the inter-electrode distance is much higher in our case, giving advantages to our cold emitter with respect to other reported work (see Table 3). The straight-line nature of the F–N plot (Fig. 10b) confirms that the electrons are emitted by the process of cold emission and the enhancement factor (β), which can be calculated from the slope (m) of the F–|N plot using the relation m = −bϕ3/2/β and the same plot gives the values of the effective work function (ϕeff) after using the relation ϕeff = ϕ/β2/3. In the calculation, the values of ϕ have been taken as 5.09 and 4.05 eV for pure and Ni-doped ZnO, respectively, following the work reported by C. Chai et al.57 The calculated values of the corresponding parameters are summarized in Table 2.
Sample | Electrode distance (μm) | E TO (V μm−1) | β | ϕ eff (eV) |
---|---|---|---|---|
ZnO | 300 | 5.34 | 2399 | 0.0299 |
Ni doped ZnO | 300 | 2.84 | 3372 | 0.0190 |
The above results clearly suggest that the doped sample has much better emission characteristics compared to pure ZnO. There are different possible reasons that alone or collectively can explain the significant enhancement of the cold emission characteristics of the doped sample.
Firstly, it is seen that due to Ni doping, there is a change in the morphology of the sample, and unlike in sample A, sample B shows the presence of a few two-dimensional sheet-like structures with sharp edges. These edges have a favorable configuration for the field lines to be concentrated into it and, thus, overall, have a higher enhancement factor. Similar results can be seen in Fig. 9, where the ANSYS-Maxwell simulation clearly shows how the field lines concentrate on sample B much more than sample A.
Defect-induced states also play an important role in monitoring the FE characteristics of a cold emitter. The XPS study shows that Ni-doped ZnO has much a larger number of defects, mainly from oxygen vacancies. In this regard, the work reported by Chen et al.58 showed that ZnO nanowires were synthesized via CVD and subsequently annealed. The as-grown sample were subjected to annealing at a temperature of 800 °C for 60 min, with a constant oxygen flow of 50 sccm. It was seen that when the cold emission characteristics of both the samples were taken, the as-grown sample gave much better emission characteristics compared to the annealed one. It was thereafter concluded that after annealing, the oxygen defect density decreased, resulting in higher resistivity in this sample, which in turn degraded the FE properties of the annealed sample considerably. Therefore, the higher oxygen vacancies in sample B can be another reason for better FE characteristics in such a sample.
The inclusion of Ni on the ZnO lattice results in favorable band bending, as discussed in-depth by Rana et al. in their reported work on the cold emission properties of Ni-doped ZnO.37 Following their explanation, the detailed energy band diagram schematically is shown in Fig. 11a–d.
Doping of ZnO with transition metals like Ni results in a shift in energy levels and an enhanced carrier population in the conduction band of ZnO. In Ni-doped ZnO, Zn2+ ions, located in the middle of the tetrahedron surrounded by four oxygen atoms, get partially replaced by Ni2+ ions. As a result, the d orbital of Ni gets split into a higher triplet (t2g) and lower doublet (eg) state (Fig. 11a) under the crystal field of ZnO. After hybridization, the t2g state interacts with the p orbital of the valence electrons, causing it to split into bonding and anti-bonding t states (Fig. 11b). The t-bonding state is localized, while the t anti-bonding state has higher energy and is positioned very close to the conduction band. This anti-bonding state has a considerable number of mobile electrons, and there is a huge probability that a slight increase in the number of electrons from this impurity state can facilitate their transition into the conduction band of ZnO. This phenomenon leads to an enhancement in the field emission (FE) characteristics of sample B (Fig. 11c and d). Thus conclusively, it can be considered that due to the incorporation of Ni, a considerable number of electrons get promoted to the conduction band of ZnO and take part in the FE.
Similar results have been theoretically verified by many researchers from DFT studies.
For instance, Haq et al. computed the density of states of Ni-doped ZnO, which is shown in Fig. 12.59
Fig. 12 The spin polarized DOS of Ni doped ZnO. The spin up and down states are respectively plotted along +Y and −Y axes. The Fermi level is represented by the dotted line at 0 energy.59 |
Fig. 12 shows that the VB at a lower energy value is mostly contributed by the Zn-3d state, whereas the middle energy range contribution comes from Zn-3d, O-2p, and slightly from the Ni-3d level. The higher energy value of the VB (−2 to −1 eV) is solely due to Ni-3d together with the O-2p level. Due to the crystal field effect of oxygen, Ni 3d states are split into eg and t2g states, as mentioned before, and contributed to the FE characteristics of sample B.
Table 3 compares the cold emission result of different ZnO nanostructure-based systems to that of our present work.
Sl no. | Sample | Synthesis method | E To at (V μm−1) | Electrode distance (μm) | Improvement [(ETo2 − ETo1)/ETo1] × 100 (%) | Ref. |
---|---|---|---|---|---|---|
1 | Pd-adsorbed ZnO | Sputtering/wet chemical | 6.4–6.6 at 10 μA cm−2 | Not disclosed | 3.12 | 34 |
2 | In doped ZnO | Thermal oxidation | 7.1 10 μA cm−2 | Not disclosed | N/A | 35 |
3 | Ni doped ZnO nanowires and nano-tips | Thermal evaporation | 3.25 for nanowires | Not disclosed | 112 | 36 |
1.53 for nanotips at 100 μA cm−2 | ||||||
4 | Sn doped ZnO tetrapod | Thermal evaporation | 3.91–1.96 at 10 μA cm−2 | 500 | 99 | 60 |
5 | Ni doped ZnO | Wet chemical | 2.5–2.3 at 1 μA cm−2 | Not disclosed | 8 | 37 |
6 | ZnO NW | MOCVD | 4.1 at 10 μA cm−2 | Not disclosed | N/A | 61 |
7 | CNT-ZnO hybrid | Solid state heating | 9.7–2.6 at 10 μA cm−2 | Not disclosed | 273 | 62 |
8 | Ge doped ZnO NW | CVD | 4.1–3.5 at 10 μA cm−2 | 300 | 17 | 63 |
9 | Cu doped ZnO QD | Wet chemical | 5.89–5.04 10 μA cm−2 | 500 | 16.8 | 64 |
10 | Al doped ZnO NRAs/graphene | CVD/sputtering/wet chemical | 1.80 V mm−1 | 170 | N/A | 65 |
Cu doped ZnO NRAs/graphene | 1.51 V mm−1 | |||||
Not defined | ||||||
11 | Ni doped ZnO NR | Wet chemical | 2.84–5.34 1 μA cm−2 | 300 | 88 | Present work |
It is clearly seen that the present work gives results comparable to or sometimes even better than other reported results with the further advantage of a large inter-electrode distance and, most importantly, higher relative improvement of the hybrid system comparable to the pure one.
Cold emission current density undergoes an order enhancement after doping, reducing the turn-on field from 5.34 to 2.84 V μm−1, with an inter-electrode distance as high as 300 μm.
The enhancement factor for the doped sample is nearly 3500 and almost 1.5 times that of the pure sample.
The enhanced emission characteristics are believed to be due to the collective results of microstructures with a high aspect ratio, enhanced number of defect-induced states, favorable band bending, and carrier flow. Thus the present work successfully reports the synthesis of a ZnO-based superhydrophobic field emitter.
This journal is © The Royal Society of Chemistry 2023 |