Oxidative and colloidal kinetics of size controlled copper structures through surface plasmon regulated examinations for broadband absorbance

Out of all the candidate plasmonic metals, copper has noteworthy optical characteristics and is also economically favourable for use. However, the stability of plasmonic copper nanomaterials against the loss of the plasmonic property is a setback. The present work is on the synthesis of oxidation-stable copper micro/nanoparticles (CuMps/NPs) at ambient conditions with chosen precursors, antioxidizing agents, polymeric capping agents and chelating ligands. The Surface Plasmon Response (SPR) response of the synthesized Cu structures and their morphological analyses are studied. The refined XRD data were subjected to a detailed structural investigation over fundamental aspects such as crystallite sizes, distortion and dislocation densities. We present herein micro/nanostructures of oxidation-stable plasmonic Cu. The validation of the aggregation and oxidation stabilities of the different synthesized samples make them a worthy choice for multiple plasmonic applications, along with showing the synthesis protocols as viable approaches for achieving such structures with a markedly increased shelf life.


Introduction
Plasmonic nanoparticles command enormous attention due to their unique thermal, electrical, and even catalytic properties apart from their optical characteristics.Plasmonic metal nanoparticles like Au, Ag, Cu, Pd, and Pt in particular are well-explored due to their high free electron density, tailorable dielectric constant, specificity for particular reactions and having their optical limit in the UV-Vis due to higher electron densities than other plasmonic materials, among other benefits.Among these candidates, copper nanoparticles (CuNPs) have high thermal (400 W/mK) [1], and electrical (59.88×10 6 S/m) conductivity, catalytic activity (due to its diverse coordination chemistry) and importantly a facile tunability in plasmonic property through uncomplicated manoeuvres such as during and post-synthesis (including controlled oxidation) and hence have potential in a wide array of applications such as in photothermal therapy [2], photocatalysis [3] and photovoltaics [4].High-temperature applications are usually excluded due to high sensitivity and susceptibility towards oxidation (the same being a cause for worry even in atmospheric temperatures and ambient).Researchers have employed numerous methods like chemical reduction [5], sol-gel [6], photochemical [7], microwave [8], reverse micelles [9], microemulsion [10], hydrothermal [11] and electrochemical [12] to synthesize and stabilize Cu particles.In the case of different oxides of Cu like Cu2O, CuO dampens the localized surface plasmon resonance (LSPR) [13] intensity of the metal core (Cu) with an associated increase in band gap and hole mobilities.The oxidation problem can be precluded by multiple means, as has been reviewed recently by our group [14,15] through the addition of ligand molecules, capping agents, or stabilizing agents viz., polyvinyl pyrrolidone (PVP) [16], cetyl trimethyl ammonium bromide (CTAB) [17], L-ascorbic acid (AA) [18], Sodium dodecyl sulphate (SDS) [19], and ligand molecules like ethylene diamine (EDA) [20].Usually, molecules which can play a dual role of reducing as well as capping are intuitively preferred due to advantages such as less intensive post-processing to isolate the CuNPs.Two mechanistic phenomena usually dictate the capping stability of such dual-role antioxidant ligands.One is the reduction/capping effect by coordinating a lone pair of electrons which can be accepted by two sp orbits of the copper ion to form a complex compound, wherein antioxidants such as ascorbic acid (AA) thus cap copper ions through the reduction of Cu 2+ inside nanoscopic templates within which small CuNPs can further grow.The second governing mechanism of oxidation stability is the dispersion effect of the oxidation product of L-ascorbic acid on the CuNPs after the completion of the reduction reaction where L-ascorbic acid is first converted into dehydroascorbic acid, and after a sequence of reactions, as will be discussed later in this study, the polyhydroxy structure is obtained whose extensive number of hydroxyl groups can facilitate the complexation of CuNPs to the molecular matrix by inter and intramolecular hydrogen bonding, preventing aggregation of Cu nanoparticles.From these detailed revealing of the synthesis process, we have explored the impact of using AA alone or in combination with two well-known capping ligands of PVP and triethanol amine (TEA) in this study, with observations of changes in particle surficial features, polydispersity and particle size as a result without a change in morphology.PVP's role as a capping polymer (albeit being a mild reductant due to its high electron affinity in a few cases like Ag, Pt) [21][22][23] to exclude aggregation, and TEA has been investigated for its relative stabilizing effect for realizing oxidant-free stable CuNPs and Copper microparticles (CuMPs).
We discuss in detail the choice of concentrations of each reagent with corroborations of the results from other studies as well as with relevant characterization tools.A few reports also include other growth manipulators such as pH-controlling reagents like NaOH, HCl as well as reducing agents like NaBH4, trisodium citrate, hydrazine hydrate, oleic acid, glucose, and so on have been employed to synthesize stable Cu nanoparticles, with and without the necessity of specific growth ambient such as N2 and H2 atmospheres.This work focuses on the possibility of preventing Cu from oxidation due to a favoured stabilizing property of the AA, PVP, and TEA combination.We also explore two precursor options of copper nitrate (CN) and sulphate (CS).The nitrate precursor, with a single anionic charge results in relatively lesser lattice enthalpies and has a high possibility of creating hydrogen bonds with water enhancing solubility by increasing the enthalpy of hydration.However, the dissociation potential of copper nitrate (strongest basic) is lower than copper sulphate (strongest acidic) indicated by (acid dissociation constant) pKa= -5.6<-3 [24,25] and the pKa value is directly proportional to the standard Gibbs free energy change [26].The effects and influence of nitrate and sulphate as precursors will be discussed in detail in the later section.The synthesis we describe is thus readily performed in air with an aqueous solution-based process exhibiting oxidation stability for a considerable period, suitable for a wide range of applications.

Material synthesis
Copper nitrate trihydrate (Cu(NO3)2.3H2O-99%purity), Copper sulphate pentahydrate (Cu(SO4).5H2O-99%purity) and triethanolamine (C6H15NO3 -99%purity) were purchased from Sigma-Aldrich.Polyvinyl pyrrolidone ((C6H9NO)n -99%purity -molecular weight of K30) was from Molychem and L-Ascorbic Acid (C6H8O6 -99% purity) was from Loba Chemie.All were used without further chemical treatment and purification and the experiments were carried out using double deionised water (Vent filter MPK01).The choice of reducing / structure directing agents and their concentrations has been validated in section 3.1, as doing so will elucidate better the original findings from this study.

Synthesis of Cu using AA
CuMPs were synthesized by adding 10 ml of aqueous copper nitrate trihydrate solution (10 mM) to an aqueous solution of ascorbic acid (0.5 M in 10 ml).The solution was stirred at 500 rpm at room temperature for 30 mins.The resulting solution turned from light blue to dark brown confirming the reduction of Cu 2+ to Cu 0 will be abbreviated as CNA.The same procedure was carried out by replacing the nitrate precursor with copper sulphate pentahydrate in which the dark brown colour was observed at a much earlier instance of 10 mins and the sample will be abbreviated as CSA.

Synthesis of Cu using PVP and AA
For this synthesis process, the same procedure and concentrations were followed to prepare CuMPs as in 2.1.1 except for the order of addition.Aqueous PVP (0.3 mM-10 ml) was mixed with the precursor (CN) and turned green, an indication of possible ligand complexation between Cu-PVP or even a mild reduction of Cu.Drop-wise addition of AA into the mixture turned it reddish-brown indicating the nucleation of Cu, with resulting samples abbreviated as CNP and CSP respectively based on the use of CuN or CuS.

Synthesis of Cu using TEA and AA
The influence of ligand molecules on particle size has been investigated and CuNPs were obtained by substituting TEA instead of PVP (2 ml-1%) in the initial step as in 2.1.2,in which a colour change to navy blue was observed.Upon the addition of AA, the colour changed from navy blue to brown indicating the formation of Cu 0 and the sample will be abridged as CNTA.Owing to the high reduction potential of sulphate, the sample has not been reduced with the same concentration of AA, details of which will be further elucidated in section 3.3.1.A holistic illustration on the preparation procedures is shown in Fig. 1.
The precipitated Cu particles from all processes were subjected to centrifugation at 8000 rpm in deionized water to acquire the final micro/nanoparticles and re-dispersed in deionized water.To obtain consistent dispersion before every characterization analysis, the precipitated samples were subjected to a 5-minute sonication.Identical quantities of all samples were taken in quartz cuvettes for UV-Vis analysis prior to a water reference being acquired and subtracted, and the concentrations were constrained to absorption conditions of less than 1.For SEM and XRD analysis, a glass substrate was washed in isopropyl alcohol for five minutes, and then sonicated in an aqueous solution.Then the samples previously dispersed in the solvent were drop-cast on the glass substrate and allowed to dry at room temperature.For the stability studies, the same substrates were stored in an ambient environment for a period of more than three months, post which the characterizations were repeated, as discussed in section.3.4.
Fig. 1 Schematic of seedless synthesis of Cu structures from nitrate and sulphate precursors.

Characterizations
The synthesized samples were characterized using UV-Vis absorbance spectroscopy with achromatic light from a tungsten halogen lamp (HL-2000 from Ocean Optics-300 nm-2000 nm) passing through the sample with a path length of 1cm.An optical fibre cable was connected from the collecting collimating lens to the spectrometer (Ocean Optics Flame-T-Vis-NIR) to get the spectra.Crystallinity was investigated by X-ray diffraction (XRD-Empyrean, Malvern Panalytical) with a Cu-Kα source, (λ=1.54Å) and the diffractograms recorded between 2θ angles of 20º and 80º with instrumental broadening correction built-in.The phase and structural corroborations were performed by Rietveld analysis associated with the Crystallographic Open Database (COD) using X'Pert High score plus by Malvern Panalytical.
Elemental composition and binding interaction information were analysed by Fourier Transform Infrared (FTIR-Shimadzu IR Affinity model-1s) spectroscopy using a double beam spectrometer wherein transmittance was collected between 4000 cm -1 and 400 cm -1 in Attenuated Total Reflection (ATR) mode.
Scanning electron microscopy (SEM) images were obtained using ZEISS EVO 18 with an operating voltage of 20 kV.Transmission electron microscopy (TEM) images and selected area electron diffraction (SAED) spectra were collected using a JEOL JEM 2100 with 200 kV accelerating voltage to study the morphological and structural aspects post-coating on copper grids and drying at room temperature.

Results and discussion
Before teasing out the significant observations from the synthesis studies, we first present the rationale for the choice of the reagents and their concentrations (section 3.1), present the generic results of characterization (section 3.2), and finally discuss the original findings (section 3.3) and the stability investigations against oxidation and agglomeration (section 3.4).

Choice of reagents
The choice of reagents has not been unexplored previously, we have carefully chosen their concentrations and sequence of addition based on predictive reasoning, as these parameters control characteristics such as stability, size, shape, intensity of absorbance spectra, and shifts in the LSPR position.
For this, a summary of the previously investigated concentrations of these reagents and the resulting particle characteristics are presented in with key observations on their influences.
From the analyses, the concentrations of the precursor were fixed as 10 mM, and the reducing agents' concentration was increased to 0.5 M (in order to reduce the particle size by the capping mechanism of AA and examine deviations to polydispersity and other features such as surface quality) and the concentration of capping agents have been decided as 0.3 mM.
As to the choice of the growth-directing agents of PVP and TEA, the choice of PVP has been validated through the observations of distinct particle structures in the table.However, from previous reports, double capping has been reported to be necessary when using specific combinations such as a different reducing agent of hydrazine hydrate for example as even though PVP acts as a growth-directing agent, a decreasing capping ability was observed compared to CTAB [36].With this in mind and since this report was aimed at realizing CuNPs with an increased temperature stability (to the tune of approx.200°C), we have explored an alternative capping agent of TEA.The amine group from TEA has a stronger surface interaction with the nanoparticles, which can result in a more stable capping action in addition to altering reaction kinetics.In summary, the particle size is expected to reduce (with an increase in capping effectiveness with temperature) when using TEA and an anisotropic growth is expected when PVP is used.Combined with the hitherto unexplored combination of AA and TEA to the best of the knowledge of the investigators, the rationale for the choice of reagents, structure directing agents and their concentrations has been decided thusly.

Results from material characterizations
The discussion on generic results of the individual characterization techniques has been given initially, post which explicit observations of the individual samples progressing from trends observed at the lattice level to the macroscopic and the aggregate level are put forth.

Structural evaluation using XRD analysis
XRD was performed on drop cast samples on a glass substrate after drying at room temperature.Fig. 2 shows the diffractograms of all samples indicating the Cu crystal phase, with a face-centred cubic (FCC) crystal system corresponding to diffraction planes of (111), ( 200) and (220).These results are consistent with ICSD no: 53757.The space group of Fm-3m with a unit cell volume of 47.24×10 6 pm 3 were confirmed for all samples, and importantly, no oxide peaks were observed from XRD results, unlike studies where this has been reported as an unavoidable issue [37][38][39].Where (k=0.9 assuming spherical particles) is the particle shape factor, λ the wavelength of the source radiation, β the full width at half maximum (FWHM) and θ the Bragg's angle.The average crystallite size was calculated as 28.2 nm, 34.2 nm, 36.8 nm, 64.5 nm, and 31.5 nm corresponding to CNA, CSA, CNP, CSP and CNTA respectively.Further insights into the finer variations between the different NPs due to the specific growth environments such as the dislocation densities, crystallinity and lattice distortions with the help of Rietveld refined data have been discussed in section 3.2.2.

Optical evaluation using UV-Vis absorption spectroscopy
The optical properties of the synthesized samples were characterized by UV-Vis absorption spectroscopy.
Fig. 3 shows the absorption profiles of Cu 0 particles, with the one around 600 nm due to absorptions by collective oscillations of plasmons.These intraband transitions can have variations in profile and position depending on particle size, shape, and distributions, the refractive index of the environment and proximity to one another, and the observed differences and the influencing causes of the same will be subsequently discussed drawing support from other characterization tools.The precursors typically undergo d-d transitions and have the absorption profile as shown in the inset of Fig. 3. Inset: UV graph for bare precursors, AA, PVP and TEA.

Morphological evaluations using SEM
Fig. 4 SEM images of prepared CuMPs synthesized from CN and CS under different conditions at the same magnifications.Insets show enlarged images for visualization of faceting/aggregation. Fig. 4 shows a clear influence on particle morphology including but not limited to faceting, aggregation, size and shape, both as a result of reactions with AA alone as well as from when PVP was added along with AA.
This gives clues on the evolution of morphology as well as on particle size distribution and aggregation with respect to other samples, wherein the particles were obtained in the range of submicron dimensions.The reactions that have led to the observed structures including that on the mechanism of PVP, reduction rate and nucleation rates of Cu over different precursors will be discussed in section 3.2.1.

Morphological evaluation using TEM analysis
For the samples prepared from TEA, particle sizes were reduced to the nanoscale and hence were analysed using TEM as shown in Fig. 5 as SEM was unable to reach the required magnifications.The average particle size (with a rough approximation of a sphere) of the nanospheres and d-spacing was calculated and found to be 82 nm ± 8 and 0.17 nm respectively.SAED patterns confirm the formation of Cu alone and d-spacing values matched well with XRD data.
From these generic discussions of the optical and structural representations, we aim to reveal a much more comprehensive picture, drawing inputs from reaction mechanisms as well as through corroborations with the experimental techniques next.

The formation mechanism of Cu 0 in different reaction environments
The reduction mechanism pathways are similar for both the precursors with the reducing agents [31,35] apart from changes to the reduction rate and the nucleation rate which will vary due to the different dissociation potentials as evidenced by the larger electrostatic attraction between Cu 2+ and SO4 2-than that between Cu 2+ and NO3 - [40].Hence, the reduction rate will be higher when CN is used as a precursor.For clarity hence, the mechanism occurring during the synthesis of CuNPs with CN as a precursor has been discussed in detail starting with the addition of the precursor to the solvent and the subsequent reactions as follows in Fig. 6.
With precursor and AA, the reaction proceeds as, Cu(NO3)2.3H2O(sol)+ C6H8O6 (sol) + H2O  Cu 0 : C6H8O6 Briefly, for the nitrate precursor, precursor dissociation into Cu 2+ and NO3 -with the NO3 -ion volatilizing as nitrogen dioxide is the common preliminary reaction step [41].The dissociated Cu 2+ ions interact with the antioxidant/stabilizing/ligand molecules which can be briefed keeping in mind the individual reactions.
a. AA donates electrons by the homolytic cleavage of an OH-bond wherein the carbonyl double bond in the lactone ring forms a conjugate with cu + accompanied by a colour change to green.The resulting version of AA (dehydroascorbic acid) has a very short lifetime time (a few milliseconds) and forms 2,3-diketogulonate [42,43].A complete reduction to Cu 0 follows and is observed by a colour change to brown [44].As mentioned earlier, a capping layer of AA molecules is expected.
b.Following the order of addition during synthesis, PVP complexes with Cu 2+ ion by donating a lone pair of electrons from oxygen atoms in the polar groups of PVP to construct metal-O2 coordination [45].
c.After the addition of AA CuNPs form and possibly the positive charge of PVP attracts to and remains as the surface charge.A similarly stable complex of CuNPs can be achieved by chelating ligands like TEA.
d.The reaction mechanism of TEA is similar to that of PVP conjugated with Cu 2+ ion but differs in terms of the properties of the functional groups that coordinate.In TEA, the process of amination (ligand molecules and especially the nitrogen atom attached to the metal atom) occurs with metal ions to form a stable complex [46].In addition, the tetradentate ligand can be formed as a bridging ligand between the alkoxy group of TEA and Cu 2+ ion [47].
e.This stable ligand complex has been reported to have good dispersion stability [48] and will reduce Cu + by accepting electrons from AA.The fact that the reduction potential of CS is higher than CN which may have prevented reduction and ligand complexation with TEA is proposed to be the reason due to the absence of CuNPs formation at room temperature.As to the absence of CuNPs formation which would have led to the CSTA sample, post complex formation of [Cu(C6H15NO3)(H2O)4].SO4.H2O [49], the weak reducing nature of AA might not have been supportive for the reduction of Cu 2+ ions.This could have led to the absence of reduction of the Cu ions into NPs/MPs.

Investigations of differences in formed lattices
With these summarizing mechanisms, we delve into the details of the lattice structure and their changes, and present calculations of pertinent parameters from the X-ray diffractograms in The interplanar spacing (d-spacing) for Cu was calculated using the Bragg equation, Where a is the lattice parameter for a face-centred cubic crystal system and hkl are the miller indices.Since the dielectric constant and the optical refractive index will be equal in all directions, we have assumed for a cubic crystal system that the distribution of dislocations is isotropic and that the dislocation density can be calculated by Where n is the number of dislocation lines per unit volume and D is the average crystallite size.The crystallite sizes of CuNPs are higher than the CNA but comparable with CSA.Combined with the observation of larger particle sizes of CNA (which compares with CSA with regards to particle size) a qualitative conclusion of a higher crystallinity (i.e., a greater number of grains) of CNA compared to CNP can be made.Since PVP and TEA are growth-directing agents [50], changes in the degree of the same can result but the growth environment for our samples has not led to a tangible observation on the same.The CSP particles are a clear outlier in the crystallinity trends, with a 2X increase in crystallite size with a simultaneously marked decrease in particle size.The dislocation density is inversely proportional to crystallite size, the sample CSP has a considerably lower dislocation density as well, exhibiting considerably larger crystallite domains.As a consequence of crystallite size, the dislocation density also follows a similar trend to the former, with CNA having the highest dislocation densities due to the lowest crystallite sizes.By refining the data from Rietveld calculations and calculating the distortion values by Where a0, a are the lattice constants of the reference lattice (from ICSD data) and the sample lattice respectively.
With minor changes to the unmodified Cu lattice and tangible differences in the crystallite sizes based on the choice of reagents, we consider next the observations at the particle / aggregate scale.

Probing particle/aggregate level observations with SEM and TEM
From Fig. 4, the particles are in general highly faceted, with differences in size distribution as well as aggregation as approximately spherically arranged structures with no specific dimensional growth apparent.Since the irregularly aggregated particles will not support calculating the particle size with accuracy, their sizes can only be concluded to be with a diameter in the range of a few microns for all the samples except CNTA.Since all samples were distributed in the same solvent and sonicated for the same durations before drop casting on the imaging substrates, it may be concluded that the images show higher clustering in CNA than in CSA.By the addition of PVP, clustering was found to be reduced with both precursors, and the presence of the organic shell (PVP) around the particles has resulted in the reduced particle size as well.From the CNP and CSP samples, it seems that the concentration of PVP was probably insufficient for it to play a role as a growth-directing agent, as no distinctive morphologies could be observed with only a reduction in particle sizes as compared to the AA-only samples.
The differences in the mechanism of capping TEA and PVP merit discussion as significant differences can be seen in their manifestation of this aspect.With TEA, the Cu + -N bond being localized to the amine group, the adsorption properties of new Cu nuclei onto existing Cu atoms will determine the size and shape of Cu 0 [51] whereas in PVP, the chain length determines the shape and surface [50] features of the particles.With regards to the TEA sample, a noteworthy phenomenon is a more defined LSPR peak.In metal nano/microparticle colloids, the polydispersity in sizes is a major driving factor for the broadening of the LSPR peak.The CuNPs of the CNTA sample, being significantly smaller in size (albeit not very small as to reach the mean free path [52] of Cu of 40 nm wherein surface damping effects start to dominate) exhibit reduced broadening of the LSPR peak due to reduced phase retardation (compared to larger particle where inhomogeneous polarization results in an increase in this phenomenon) and reduced multipole excitations [53], as could be observed in the absorption spectra shown in Fig. 3.
The testing of the samples towards their oxidation, as well as aggregation tendencies with temperature and time on a substrate when left exposed to an air ambient, revealed unusual results, as will be discussed and validated with relevant characterization next.Specifically, the changes were examined in the context of absorption spectra as well as through crystallinity studies.Although the imaging of the same particle clusters before and after these studies would be ideal through an additional microscopic comparison to the crystallinity and absorption studies, the requirement of sputtering a conductive layer of Au over the particles before SEM imaging prevented this study, as the presence of Au can significantly alter the results of the investigations.

Discussions of the stabilities of the CuMPs/NPs
It is better to be defined at the outset that the stability in this work stability has been taken to mean the tendency of the particles against two aspects viz., aggregation and oxidation.In this regard, stability depends on the dispersant layer, pH, temperature, surface charge of the nanoparticles and so on.The colloidal stability of the CuMPs/NPs preserves the size, and shape of the synthesized structures and preventing from agglomeration.Cu nanoparticle oxidation can be passivated via polymer-capped or ligand-functionalized techniques that encapsulate the metal core with surface charges without electrically insulating it.The stability of nanoparticles influences the LSPR position and spectral width by its particle size distribution.However, stable Cu nanoparticles play a major role in a wide range of applications like water desalination, photothermal conversion and other plasmonic applications.Consequently, the next part will expand on the detailed discussion regarding the stability of oxidation and aggregation.

Aggregation stability
Surface charges imparted during synthesis from dispersing/reducing/capping agents play a major role in nanoparticle stability, in addition to particle density during storage and handling post synthesis.The shape changes induced by electrostatic interaction between particles are favoured due to Ostwald ripening, which occur during storage or prolonged use of post-synthesized samples in conditions facilitating inter-particle collisions, [54] wherein the interaction pair potentials might exceed the kinetic energy barrier which prevents aggregation.This interaction pair potential is the product of the osmotic, elastic, van der Waals, and electrostatic-interaction potentials.
Examination of the trends on particle aggregation revealed interesting results, reflected in the plasmon bands shown in Fig. 7. AA/TEA show an apparent increase in the resonant bandwidth, and this increase in spectral width of nanostructures may be an indication of particle aggregation or clustering.The contribution of higher-order phase retarded plasmon modes which depend intricately on the particle: incident wavelength ratios are responsible for the asymmetric character and broadening of SPR [55].It is of note that all samples have retained their metallic nature, confirmed by the strong observance of the dipole mode at approx.590 nm along with the absence of oxide peaks (discussed in the subsequent section).It is well-established from theory and research that the extrinsic size effects of Cu structures (sized above the m.f. p-value of 20 nm) dictate that with the size the dipolar transition decreases and higher-order contributions dominate, depending to various degrees on the resulting particles' surface quality apart from the size and shape [56][57][58][59].Sonication before UV acquisition after the aggregation stability testing did not influence the peak profiles as can be seen from the Fig. 7.Only the CNA sample seems to be an outlier here, in that the absorbance profiles changed to a qualitatively much lesser degree than for the other samples, reflecting a stronger stabilizing action against aggregation when only AA is used.All of the samples, though experiencing aggregation, also had excellent and unconventional oxidation stability throughout the same duration and will be discussed subsequently.

Oxidation stability
Copper has an extremely high tendency for oxidation, and the predominant result is the oxidation product of Cu2O [60] and CuO (with intermediate phases such as CuOx where x = 0.67) with a range of hydroxides being possible.Multiple studies propose oxidation progressing as the uniform passivation of the CuNPs' surface to form the oxide [61], augmented later by the observation of oxide islands nucleating on the surface and coalescing to form a film [62].The end result is almost always a quick oxidation of pristine Cu with a resulting self-limiting layer of ~ 15 nm (composed of Cu2O) thickness, simultaneously observed by the presence/conversion of other species such as CuO and CuOxH2O.Hence it stands that oxidation while considering Cu can have significant differences in terms of the time required, and that the extent of selflimitation depends on factors purity, morphology, temperature, and surface characteristics (such as faceting wherein oxidation in Cu (111) is due to surface stress [63] and anisotropic growth where applicable, roughness, porosity, local strains etc.) and the ambient environment (with its own controlling parameters) [64].That Cu oxides cannot be avoided unless stabilized or isolated against even the mildest of oxidizing environments [65] is the conclusion.We report in this work probably one of the few observations of CuNPs/MPs stabilized significantly against oxidation for more than three months when stored in an ambient environment of air, drawing these conclusions from crystallinity studies Fig. 8 and absorption analyses Fig. 7. Similar signatures of all planes hence demonstrate the long-term stability.As for the differences in absolute intensities, it is possible that the data collection over the samples during characterization may not have been at the same position as during the collection before the start of the oxidation study.A more sensitive change towards oxidized Cu from pristine Cu is possible through UV analyses, which from Fig. 7 can also be seen to be avoided.These results elucidate that the reagents used and their concentrations could have played a significant role in preventing oxidation.

Conclusions
In this study, we report concentrations of a specific set of reagents of AA/PVP/TEA for the synthesis of oxidation-stable CuMPs/NPs.We detail the reaction mechanisms that have led to the observations for better validation with experiments.The increased concentration of AA (>0.4M) results in a significantly better stabilized Cu particle, with the effect of additional reagents such as PVP being on achieving a reduced particle size.On the other hand, PVP seems to have detrimentally (from a plasmonic point of view) caused the development of Cu microstructures and irregular aggregates.The differences in achieving particle sizes in the nanoscale vs the microscale ranges using tailored combinations of AA, PVP and TEA have also been highlighted.We are currently examining the stability of the Cu phases against oxidation when subjected to higher-than-ambient temperatures, as CuMPs/NPs exposed to air under atmospheric conditions have been confirmed to not suffer from oxidation.XRD results clarify that the pure Cu phase was not degraded even after three months due to the stabilizing nature of the reagents, making these structures ideal for achieving longer shelf lives.Results from Rietveld refinement point to considerable differences in the growth mechanism with PVP, wherein an unusual increase of the crystallite size was observed compared to the other two combinations of AA alone or AA+TEA.Tailoring such oxidation-resistant Cu structures with a stable SPR response is a significant outcome as such particles can be of interest in applications such as catalysis, photothermal conversion, and photovoltaics.

Fig. 2
Fig. 2 XRD patterns of Cu structures synthesized from nitrate and sulphate precursors with and without structure directing agents.

Fig. 3
Fig. 3 Absorbance spectra of Cu structures synthesized as a function of different precursors / capping agents.

Fig. 5
Fig. 5 TEM images of prepared CuNPs synthesized using TEA. a -d at different magnification scales, e SAED pattern for the CuNPs.

Fig. 6
Fig. 6 Schematic representation of the reaction mechanism of Cu structures from different capping/ reducing agents with multiple intermediate steps.

Fig. 7
Fig. 7 Comparative absorbance spectra of Cu structures between post-synthesis and after a period of 3 months.a. without structure directing agents.b.With structure-directing agents PVP and TEA

Fig. 8
Fig. 8 XRD patterns of Cu structures synthesized from nitrate and sulphate precursors a. without structure directing agents.b.With structure-directing agents PVP and TEA showed similar diffraction peaks of unoxidised Cu after three months.From Fig. 8 the diffraction patterns of all the samples reveal the absence of signatures from oxide phases.

Table . 1
Table. 1. Summary of concentration investigations of AA/PVP on the morphology of Cu nano/microparticles

Table . 2
Table. 2. Analyses of the crystallite sizes, dislocation densities, lattice distortions and d-spacing of individual planes for all the samples.