Surface elemental and structural analysis of ancient Indian punch-marked coins (600 to 200 BCE): insights into metallurgy and economic practices

Abstract

Punch-marked coins (PMCs) are the oldest coins in India and among the most widely circulated globally, often found in hoards that highlight their extensive use. This study utilizes X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) to analyze the surface elemental composition and chemical properties of nine series (S-0 to S-VIII) of Janapada (S-0) and imperial PMCs (S-1 to S-VIII) dating from 600 to 200 BCE, housed in the Numismatic Society of India at BHU, Varanasi, based on the Gupta-Hardakar classification related to the PMCs. XRD results reveal four prominent diffraction peaks corresponding to metallic silver (Ag) in the face-centred cubic (fcc) phase, with a slight variation in d-spacing (∼ 0.05 Å), suggesting subtle changes in the lattice structure due to smaller atomic radius elements. XPS analysis shows the non-uniform distribution of different elements, with Ag being predominant, alongside copper (Cu), lead (Pb), and trace elements, including gold (Au), only in Janapada PMCs (S-0). The binding energy curves indicate that Ag and Cu are in pure metallic forms, while Pb exists as Pb₂O₃. Importantly, no silver or copper oxide peaks were detected, indicating the metals’ purity throughout the coinage process. The variations in d-spacing observed in these historical samples offer a microscopic perspective into the broader contexts of ancient economies, technologies, and cultural practices. The silver content in these PMCs decreases as Cu and Pb increase across the series up to S-V, followed by a sudden rise in S-VI, which lacks Pb. The presence of Pb induces brittleness and may serve as an indicator of the coins’ age. Additionally, the carbon detected by XPS could result from smelting, surface contamination, or environmental deposition. These findings reflect a high level of metallurgical knowledge and alloying techniques developed between the sixth and third centuries BCE. The varying Ag content raises questions about the economy and demand for coins. At the same time, the structural variations identified through XRD and XPS can aid archaeometallurgists in estimating these coins’ chronological and geographical origins, serving as valuable tools for authentication.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-76356-3.

Introduction

Recently, with the advancement in various spectroscopic and microscopic techniques, the study of historically important punched marked coins become very fascinating for getting various chemical compositions at the microscopic scale^1–10. Analyzing the chemical composition of ancient coins provides valuable insights to archaeologists and numismatists, revealing details about manufacturing techniques, age, economic status, minting locations, and authenticity. These coins, made from various metal alloys across different periods in Indian history, reflect the evolution of metallurgy. They also offer facts about ore origins and production methods^11,12.

Focusing on Indian Punch-Marked Coins (PMCs), particularly silver PMCs abundant in Asia and often found in hoards, presents complexities in classification, such as distinguishing between local and imperial varieties¹³. The transition from local to imperial PMCs coincided with Magadha’s rise as a dominant imperial power around 600 BCE, leading to significant changes in the region’s currency system. New coins merged old punch-marked styles with new imperial symbols, minted using a multiple-die striking method. This evolution can be traced through distinct phases identified by Gupta and Hardaker¹⁴, reflecting advancements in design and size, and shaping a uniform monetary system under imperial rule. Studies on PMCs delve into their metal compositions, reflecting ancient economic theories like those in Kautilya’s Arthashastra¹⁵ from the third century BCE. Kautilya proposed combining metals like copper and silver to create different values of coins. P. L. Gupta’s 1963¹⁶ research aimed to ascertain copper-silver ratios in coins, finding copper content typically at 20–30% but occasionally up to 30–50%. Ahmad and Prasad¹⁷ expanded on this by exploring additional alloying elements like tin, iron, lead, and antimony as hardening elements. Various studies of the analysis of the chemical composition on the surface of the ancient coins by nondestructive X-ray fluorescence (XRF) techniques were reported in the literature^18–22. Based on XRF techniques, Vijayan et al.³ compare the Indian PMCs and Alexander’s coins revealing differences in silver element dominance in Alexander’s coins. Meenakshi et al.²³ also used XRF to distinguish between genuine and fake medieval Indian silver coins. Detection of Ag and Cu on the surface by µXRF techniques shows that emphasizing the peak intensity ratios of Ag Kα/Lα, Cu Kα/Ag Kα, and Cu Lα/Ag Lα between coins suggests that employing a multi-standard approach offers the most dependable identification of Ag enrichment and Cu depletion on the surface^24,25, The analysis of PMCs by particle-induced µX-ray emission (PIXE) techniques identifying a range of trace and minor elements including K, Ca, Ti, V, Cr, Mn, Co, Ni, and Rb, alongside major ones like Ag, Cu, Au, Pb and Fe enhancing our grasp of ancient coin metallurgy and economic practices. Mamania’s²⁶ team used XRD and SEM-EDX analysis to determine that eight silver coins were made of a silver-copper alloy, extracted from argentiferous galena through cupellation. Marussi et al.²⁷ examined 160 denarii and antoniniani using µ-EDXRF and SEM-EDX to assess composition changes and currency devaluation in Roman coins. Volpi et al.¹ explored medieval denarii characteristics using p-XRF, SEM-EDX, and FTIR, highlighting the complexities of analyzing diverse data from archaeological artefacts like coins, which exhibit significant heterogeneity¹. These investigations illuminate ancient minting practices and the use of metal alloys to meet coinage demands.

In this study, we analyze variations in copper and silver across nine-coin specimens, minted between 600 and 200 BCE, selected based on Gupta and Hardaker’s classification¹⁴. Unlike previous research, which often randomly selected coins^1–24, our approach is more structured, considering both the historical and economic contexts of the coinage. While past studies primarily employed surface analysis techniques such as XRF, µXRF, PIXE, and SEM-EDX, they lacked a comprehensive exploration of the broader technological and economic implications of these artifacts. Using two advanced x-ray spectroscopic methods, we examine surface elemental compositions across various coin series, comparing imperial coins with those from the Kashi Janapada (S-0). This approach offers a deeper understanding of coinage across diverse temporal and political contexts. Key aspects of our research include the dominance of silver in coin compositions, its gradual decline over time, variations in the silver-to-copper ratios, and the notable discovery of gold in Janapada PMC. Additionally, we explore the economic consequences of coin debasement during this period and the advancements in metallurgical techniques. By integrating Gupta and Hardaker’s manual classification, which is based on the presence of symbols on PMCs, our research takes a scientific approach to unveil insights into the coins’ composition, production methods, and historical significance. Our work aims to fill the gap by providing a more methodical numismatic framework, leading to a clearer understanding of the metallurgical knowledge of the time, and its impact on the economy and currency standards. Scheme 1 illustrate the fundamental concepts of XRD and XPS as applied to examining PMCs, highlighting the presence of various elements. The research findings in future may address crucial aspects like finding spots, and metal sources for PMC research, to enhance our understanding of these coins in India’s ancient monetary system.

Scheme 1.

PMC Exposed to X-Ray: Element Composition Analysis Using XRD and XPS Techniques.

Historical background of the PMC

Punch-marked coins (PMCs) stand as India’s earliest coinage, primarily in silver or silver-based alloys, occasionally in copper. These coins differ in forms, sizes, and weights and are thought to have circulated from the seventh to first centuries BCE, extending into the second century CE post-circulation. They are categorized into three main groups: Janapada coins (600 BCE – 400 BCE), Imperial Coins (500 BCE – 200 BCE), and Provincial Coins (200 BCE – 100 BCE), with Imperial coins further classified by size (large, medium, and small). Gupta-Hardakar’s²⁸ classification system further refines imperial coins into Series I (large), Series II, III, IV (medium), and Series V, VI, VII (small). Additional series include Series 0 (pre-Series I) and Series VIII (post-Series VII), reflecting regional and imperial transitions during the Mauryan period. Imperial coins maintained a consistent weight standard of around 3.4 g known as Karshapana²⁹.

The categorization of PMC into different series is not only determined by their sizes but also by the presence or absence of specific symbols. Janapada coins, also referred to as local coins, were minted locally by various Janapadas until the fourth century BCE, long after its issue had ceased, featuring one to four symbols¹⁴. These coins exhibit variability in style, weight, and fabric, making them rare numismatic finds. In contrast, Karashapana coins, or imperial coins, bear five official punches on their obverse side. These symbols are often haphazardly punched, with many coins displaying symbols that are partially out of the flan or overlapped. Gupta and Hardakar¹⁴ noted that imperial or universal PMC typically feature two prevalent symbols: the Sun and the Six-armed motif. The remaining three symbols can then be further classified into various classes, groups, and varieties, which are referred to as the third, fourth, and fifth symbols. The frequency with which the symbol depicts is the main criterion in determining which is third, fourth or fifth. Thus, the third remains constant while the fourth varies, or the fourth may remain constant while the fifth varies. Given this manual classification, in this paper, an attempt has been made to analyze the presence of surface element composition in each series from 0 to VIII. As a result, we chose nine coins. It is also high time that, based on metal composition analysis, we create a group of coins/clusters issued in the same composition, which may shed light on the available mints and minting technology at the time.

Imperial PMCs exhibit diverse shapes such as circular, elliptical, rectangular, or square, reflecting their minting techniques^30,31. The metals were first melted in crucibles and cleansed with alkalies before being beaten into sheets on an anvil with a hammer. The sheets were then cut into pieces with clippers to a specific weight³². The irregular shapes result from cutting edges, ensuring a balanced weight like 56 grains. In contrast, circular or elliptical shapes are formed through a droplet method where molten metal conforms naturally to weight standards during pouring^30,31. Based on shape, some coins were made by cutting the edge technique (S-I, S-II, S-IV, S-VII and S-VIII), while others were droplet-formed (S-III, S-V, S-VI, S-0). The finding of PMC coins is also very interesting. They have been discovered across a vast region stretching from the Oxus to the Bay of Bengal. The coins being examined belong to this geographical zone, yet their exact provenance is unknown. The coins under study are kept in the Numismatic Society of India (NSI) collection at Banaras Hindu University in Varanasi, India, where they are stored in a secure environment within metal cabinets and placed in acid-free envelops to ensure their preservation, however, nothing is known about how they were acquired. Altogether forty-nine coins are kept in the NSI collection³³. Most of the coins are quite corroded or worn out in condition, showing its longer circulation. Out of this, we have selected nine coins, each from one series for our study. The image and major find spots of nine PMCs of 600 to 200 BCE under examination are shown in Scheme 2.

Scheme 2.

Major find-spots of imperial PMCs and image of nine coins under examination.

Methodology

PMC sample cleaning and physical parameter

Nine punch-marked coins selected based on the Gupta-Hardakar classification system from 600 BCE to 200 BCE period were suitably cleaned before analysis. The coins were cleaned gently with a brush to remove the dirt and then subjected to ultrasonic cleaning (a non-invasive method) in distilled water for two days. Then, the coins were gently wiped with tissue paper for XPS and XRD measurements. Weight, diameter, Dynasty, observed sculpture and pictures taken after cleaning for different PMCs are given in supporting Table 1.

X-ray diffraction measurements

X-ray diffraction (XRD), a noninvasive method, was employed to analyze the crystallographic structure, and physical attributes of the PMC by leveraging the constructive interference of monochromatic X-rays interacting with the surface and interface layers of the PMC. The phase composition and crystallinity of the historical samples were characterized using a Miniflex 600 (Rigaku, Japan) diffractometer and Cu Kα (1.5406 Å) radiation (with an operating voltage of 40 KV and 15 mA current). This diffractometer is equipped with a fast D/teX Ultra detector. The step size of 0.02 degree, 2θ range between 20° to 80° with a scan speed of 5° per minute was used for every measurement. XRD data was analyzed using PDXL software from Rigaku, Japan.

X-ray photoelectron spectroscopy (XPS) measurements

Further, to identify the surface elemental composition of the PMC coins, along with their chemical states, electronic structure, and the density of electronic states, XPS measurements were performed. The analysis was conducted using the Thermo Fisher Scientific K-Alpha Plus, an XPS instrument³⁸ specifically designed for electron spectroscopy and chemical analysis. Thermo Fisher Scientific K-Alpha plus XPS equipment is equipped with a Monochromatic Al Kα (1486.67 eV) X-ray source, having power 72 W, used for excitation purposes. EX06 Ion gun having energy range of 200 eV to 4 KeV, used for excitation. This Thermo Fisher Scientific K-Alpha plus XPS is equipped 180° double-focusing hemispherical analyzer (128-channel detector) and dual beam charge neutralization. The analyzer axis to the sample surface is 90 degrees, Acceptance angle is ± 30 degrees (30-degree cone). The spectrometer work function is 4.2 eV and the analysis chamber base pressure is 10⁻⁹ mBar. All measurements were performed at room temperature. The spot size is 400 μm and the analyzed area of the sample is 400 μm aligned with the spot size. The survey spectrum was recorded with a pass energy of 200 eV and 1 eV per step, and high-resolution spectra were recorded with a pass energy of 50 eV and 0.1 eV per step. The accuracy of the Binding Energy (BE) value in the Thermo Fisher Scientific K-Alpha Plus XPS machine is typically within ± 0.1 eV. It is again a non-destructive surface-sensitive quantitative spectroscopic technique based on the photoelectric effect obtained by irradiating a material with a beam of X-rays. These measurements of PMC were performed under ultra-high vacuum conditions and used in line profiling of the elemental composition across the surface, or in-depth profiling when paired with ion-beam etching. The Thermo Avantage software version 5.9931 is used for data acquisition, processing, and analysis in the K-Alpha Plus XPS machine. This software provides advanced tools for peak fitting, background subtraction, and elemental analysis. The binding energy scale was calibrated using standard samples of Gold, Silver and Copper. These metals have well-documented BE for specific core levels as Au ⁴f₇/₂ is ~ 84.0 eV, Ag ³d₅/₂ ~368.2 eV and Cu ²p₃/₂ ~932.7 eV³⁶.

Results and discussion

A systematic XRD, and XPS spectral analysis of the nine PMCs stored in the Numismatic Society of India at BHU, Varanasi under different instrumental approaches is given below.

XRD data analysis of nine PMCs

The XRD data of nine punch-marked coins is depicted in Fig. 1, revealing four prominent peaks between 35 and 80 degrees of diffraction. We examined the X-ray diffraction patterns to comprehend each silver coin’s crystalline structure and elemental composition. For Sample S-0, significant diffraction peaks were noted at 37.06°, 43.22°, 64.36°, and 76.58° (DB Card Number − 9013051). In contrast, for coins S-I to S-VIII, these peaks showed slight shifts to 38.14°, 44.32°, 64.48°, and 77.42°. Shift in two theta values depends on co-constituents of silver in each coin, manufacturing method and sample geometry³⁴. Considering that silver (Ag), gold (Au), and copper (Cu) metals, along with their oxides, are typically found in PMCs, the XRD patterns were compared with reference patterns from the JCPDS database for Ag (JCPDS File No. 04-0783), Au (JCPDS File No. 04-0784), Cu (JCPDS File No. 04-0836), Ag₂O (JCPDS File No. 43–0997), and Cu2O (JCPDS File No. 05-0667). The analysis suggests that silver predominates as the main component of these coins.

Fig. 1.

XRD data of nine punch-marked coins and JCPDS reference files for Ag, Au, and Cu.

To gain deeper insights into the characteristics of diffraction peaks, the XRD pattern of all nine PMCs samples was segmented into four regions, and the magnified overlapped XRD peaks are presented in Fig. 2. Notably, a slight shift in peak positions is observed across all planes. This shift, accompanied by the broadening of the peaks, suggests the oxidation of certain surface elements. Examining the magnified diffraction peak of the (111) plane [Fig. 2(a)], it’s apparent that the peaks exhibit sharpness (Lorenzen shape), albeit not precisely overlapped, with variations in intensity observed among these planes. Additionally, a discernible hump is evident in the peaks of S-II and S-VI at angles 38.46° and 38.62° of 2θ, respectively. In the magnified diffraction plane (200), overlapping is observed between S-I and S-VIII, S-II and S-IV, and S-III and S-VII PMC samples, accompanied by variations in intensity. Likewise, overlapped magnified peaks in planes (220) and (311) exhibit shifting in peak positions and broadening, particularly in S-II, S-V, and S-VI PMC samples, as depicted in Fig. 2(c) and 2(d), respectively. Small overlapped peaks appear at angles 65.1° and 78.0° in planes (220) and (311), respectively, alongside the main peaks. These variations in XRD diffracted peaks suggest the presence of different elements as impurities or the oxidation of surface elements of PMC samples. Furthermore, the decrease in peak intensities indicates a reduction in the quantities of major silver (Ag) components in these coins.

Fig. 2.

Magnified Overlapped peaks of nine punch-marked coins (a) for peak (111) (b) for peak (200), (c) for peak (220) and (d) for peak (311).

Compared to the pure silver XRD pattern, a slight shift in peak positions is observed in all nine samples. This shift in XRD peak position suggests possible alloying or mixing of elements. The variation in 2 theta values, within 0.42° for coins S-I to S-VIII, depends on the co-constituents of silver in each PMC, the manufacturing method, and sample geometry. Similar peak position shifts of silver with minor copper content were observed by Tsuji et al.³⁵. These peaks were identified as corresponding to the (111), (200), (220), and (311) planes of the face-centred cubic (fcc) phase structure of silver with lattice constants a = b = c = 4.086 Å, and α = β = γ = 90°, belonging to space group Fm3m¹. It’s worth noting that the lattice constants for silver and gold are similar, making distinguishing their diffraction peak positions challenging. The presence of copper and gold in the coins is in very small concentrations, which would not significantly affect the lattice constants. In coins S-0, S-IV, S-V, and S-VI, the Full Width at Half Maximum (FWHM) of the peak around ~ 65° is large, with a small shoulder present. This shoulder could potentially be attributed to the presence of copper. However, due to the lower concentration of copper, it’s challenging to definitively assign it. The presence of sharp and intense diffraction patterns confirms the crystalline nature of the coins. Additionally, no diffraction peaks corresponding to silver oxide or copper oxide were observed, indicating that silver and copper are present in pure metallic form. This purity preservation is crucial, as oxidation can alter the physical appearance and the structural properties of coins, making them less valuable or more brittle over time.

Further, slight variations in d-spacing across different series of coins (S-0 to S-VIII) were observed, specifically for the (111) plane, with values of 2.40, 2.33, 2.36, 2.35, 2.36, 2.33, 2.34, 2.35, and 2.35 Å, respectively. The overall variation in d-spacing is approximately 0.05 Å from S-0 to S-VIII. This variation can be attributed to the presence of smaller atomic radius elements, such as Cu, Al etc. within the silver matrix, whether from the ore or the mixing of elements during the coin production process. Despite using the same elements, differences in temperature and duration during various historical periods of coin striking can lead to variations in d-spacing. These discrepancies may arise from the intermixing of elements, resulting in distinct d-spacing in coins from different eras. Factors such as the availability of raw material’s desired physical properties, and changes in minting technology over time can influence these variations. Consequently, the conditions under which elements are mixed including temperature and duration can significantly affect the characteristics of the resulting coins. The presence of smaller atomic radius elements compresses the lattice structure, reducing unit cell dimensions and causing shifts toward higher diffraction angles and a change in the atomic arrangements within the metallic structure of PMCs. Higher temperatures can enhance atomic mobility, promoting a more homogeneous mixing of elements, while lower temperatures may lead to incomplete alloying and the retention of distinct microstructures. The variations in d-spacing observed in the historical coin samples may provide a microscopic window into the macroscopic world of ancient economies, technologies, and cultural practices by offering detailed insights into the alloying techniques, minting conditions, and economic priorities of different historical periods. By conducting in-depth structural analyses, researchers can trace the evolution of coin production methods and better understand the political and economic contexts in which these coins were created.

These variations in d-spacing can also serve as a valuable resource for archaeometallurgists when linked to established historical timelines and minting practices. By comparing d-spacing values from newly discovered coins to those of historical samples, researchers can estimate the likely period and geographic origin of the coins. This approach also aids in authentication, as coins exhibiting unusual d-spacing values relative to their claimed time period may be flagged as modern forgeries or reproductions, given that contemporary metallurgical processes typically yield minimal d-spacing variations and more uniform crystal structures, reflecting advanced technological capabilities.

XPS data analysis of nine PMCs

Further, to understand the shift and broadening of XRD diffraction peaks, XPS measurements of the surface of the nine PMCs were carried out to understand the presence of different elements’ composition and their chemical bindings³⁶. A relative percentage of elemental composition at the surface of the different series of PMCs and a local Kashi coin obtained from XPS measurements are shown in Table 1. Silver (Ag), copper (Cu), lead (Pb), sulphur (S), chlorine (Cl), aluminium (Al), calcium (Ca), manganese (Mn), carbon (C), oxygen (O), nitrogen (N), and silicon (Si) appears in different percentage in S-I to S-VIII PMCs, whereas Gold (Au), is only present in S-0 Kashi coin along with other elements. Variation of elementary chemical composition in a series of PMCs shown in bar diagram Fig. 3.

Table 1.

Relative percentage of elements present in different samples.

Sample	Elements relative (%)
Coin	Ag	Cu	Pb	S	Cl	Al	Ca	Mn	Au	C	O	N	Si
S-0	67.3	5.3	0.1	1.3	1.1	-	-	-	3.3	11.0	8.3	1.3	0.5
S-I	62.8	5.0	3.2	-	-	0.3	-	-	-	16.4	10.1	1.0	0.4
S-II	58.3	7.0	3.0	1.5	0.7	0.8	-	-	-	15.1	12.5	0.7	0.4
S-III	51.0	9.2	1.5	1.4	0.5	1.0	-	-	-	18.3	15.1	1.5	0.4
S-IV	49.5	7.5	1.6	1.6	0.8	0.9	0.1	-	-	19.7	15.4	1.0	1.7
S-V	38.0	11.5	0.2	1.0	0.5	1.3	0.2	1.5	-	24.0	17.1	3.1	1.7
S-VI	66.2	7.7	-	0.7	1.0	-	-	-	-	13.3	10.7	-	0.2
S-VII	37.1	6.0	2.3	0.8	1.2	0.4	-	-	-	26.3	21.0	3.8	0.7
S-VIII	55.6	6.8	2.1	0.3	0.9	0.8	-	-	-	20.0	13.3	-	-

Fig. 3.

Bar diagram of a variation of elementary chemical composition in a series of PMCs.

XPS results indicate that silver is the major component of all PMCs. The percentage of silver decreases from coin S-0 to S-V then increases maximum suddenly in coin number S-VI and further, it decreases for S-VII and increases for S-VIII. Overall, among all the samples, S-0 contains a maximum of 67.3% of silver. Copper is another constituent of the PMCs, the copper concentration in composition was 5.3% for S-0 then it slightly decreased to 5.0% for S-I. From S-I to S-III copper concentration increases from 5 to 9.2% and falls to 7.5%. The maximum 11.5% content of copper was found for sample S-5. Further for S-VI, S-VII and S-VIII change is not uniform and is found to be 7.7%, 6.0–6.8% respectively. The Variation of silver and copper percentages and the variation of silver percentage with copper percentage at the surface of the PMC coins are shown in Fig. 4(a) and (b) respectively. The results indicate variation in the moulding period. Cu mixed with Ag to slightly harden the alloy³⁵.

Fig. 4.

(a) Variation of silver and copper percentage with different PMC coins. (b) A plot of silver percentage versus copper percentage for different PMC coins.

Further, the presence of Pb in the PMC series may indicate varying metallurgical practices and technological capabilities across different historical periods. In Table 1, Pb appears in varying concentrations in most of the coin series, with the relative percentage changing significantly between the samples. In the early series PMC S-0, the Pb concentration is minimal at just 0.1%. This may suggest that during the initial phases of production, Pb might not have been deliberately alloyed and could be present as a trace impurity; however, as we progress to later series of PMC like S-I (3.2%) and S-II (3.0%), the Pb content increases, indicating a possible intentional inclusion to alter the physical properties of the coins. As a soft metal, Pb can make Ag-Cu alloys easier to cast and shape, which might have been advantageous for minting processes. This trend continues through a series of PMCs like S-III (1.5%) and S-IV (1.6%), reflecting a systematic use of lead as the minting techniques evolved. Interestingly, by series S-V, the lead content drops drastically to 0.2%, and it is absent in series S-VI. This reduction could suggest a shift in metallurgical practices, possibly driven by changes in resource availability or technological advancements that rendered the use of Pb unnecessary. The absence of Pb in coin S-VI may also indicate that this PMC was minted from natural silver ores processed without lead. The reappearance of lead in series S-VII (2.3%) and S-VIII (2.1%) indicates a renewed use of Pb, possibly due to economic factors or a need to maintain certain physical characteristics of the coinage, such as weight and appearance. Overall, the varying lead content across the PMC series highlights a nuanced understanding of alloying techniques and metallurgical control. Higher concentrations in certain series may indicate its role in enhancing castability, while fluctuations could correlate with changing political and economic conditions influencing coin production methods. In addition, surface enrichment of Pb also gives the impression of intentional lead addition to modify the alloy’s properties and increase brittleness. But it may be a natural consequence of metallurgical processes and environmental exposure. However, our XPS data shows varying lead percentages across different PMC series and the complete absence of lead in the coin series S-VI. Therefore, it is also possible that the lead content is attributed to the specific ore composition used rather than being a consequence of environmental contamination. In XPS measurements, the presence of Pb, combined with minor elements such as S and O, can form intermetallic compounds like PbS or Pb₂O₃, which may exacerbate brittleness over time. The low lead levels in other coins are possibly from the cupellation process or the smelting of argentiferous lead ores like galena. Higher lead levels could indicate intentional addition for casting, or they may result from differential corrosion. Pb in PMC coins is one of the reasons for increased brittleness; with increased concentration, it may be used to determine the coins’ age. Higher lead content is more prone to cracking and structural degradation, reflecting centuries of metallurgical changes and environmental interactions. As lead content increases, the coins become more susceptible to embrittlement.

Additionally, the large quantity of carbon detected by XPS on the surface of the PMC is shown in Table 1, which may be closely linked to historical metallurgical processes, environmental exposure, and modern contamination. High carbon content on the surface could suggest contamination and may also be a valuable indicator of the historical production environment and techniques. Using charcoal in metallurgical processes, organic moulds, or protective coatings aligns well with documented ancient practices¹⁵, making carbon a potential marker of authenticity and historical context. Ancient coins, especially those buried or exposed to atmospheric conditions for long periods, are prone to surface adsorption of carbon-based compounds. Organic materials in the soil, decaying plant matter, and microorganisms can deposit carbon on the surface over centuries, leading to a high carbon signal in XPS analysis. Calcium was present in S-IV and S-V samples in small quantities viz. 0.1% and 0.2% respectively. The presence of manganese (Mn) was observed only for the S-V sample. It is important to mention that gold was observed only for sample S-0 through the XPS measurement. The binding energy of Ag, Cu, Pd and Au observed on the surface of PMC are given in Table 2.

Table 2.

The binding energy of different elements appeared at the surface of the PMCs.

Coin No.	Elements
	Ag metal	Cu metal	Pb metal	Metal sulfide/ Thio bound to gold	Al Oxide	Metal Chloride	Au metal Pure /sulfur-bound gold
S-0 S-I S-II S-III S-IV S-V S-VI S-VII S-VIII	368.29 368.33 368.29 368.30 368.28 368.33 368.27 368.33 368.32	932.78 932.67 932.70 932.78 932.72 932.66 932.59 932.61 932.68	- 138.65 138.56 138.59 138.51 138.56 - 138.58 138.67	161.34/162.48 - - - 161.5 161.97 - - -	- - - - - 76.88 - - -	- - - - - - - 198.04 -	84.17/87.98 - - - - - - -

The overlapped and normalized binding energy XPS curves of some major elements Ag, Cu and Pb for all the nine PMC coins are shown in Fig. 5. The binding energy is inferred from the kinetic energy of the ejected electrons and the energy of the incident photons. The XPS spectra of nine PMCs have been recorded in different ranges of binding energy as shown in supporting Figures S1 to S9. Due to spin-orbit coupling, most of the elements show splitting in the binding energy and show their doublets. The slight change in the binding energy of different elements indicates, a change in the oxidation state of the surface elements. The binding energy curve corresponding to the doublet of the silver has a sharp structure. The overlapped spectra of the binding energy of Ag, in all the nine PMCs, as given in Fig. 5(a) show slight variation in the intensity of the doublet structure at 374. 28 eV and 368.29 eV corresponding to the Ag³d_3/2 and Ag³d_5/2 states corresponding to variation in concentrations of Ag in these coins as given in Table 1. The peak shape is identical in all the PMCs. These peaks match well with the Ag metal XPS peaks which indicates the presence of Ag in metallic form in S0 coin³⁷. The normalised spectra of Ag as shown in Fig. 5b show exact overlap in the doublet structure with no change in the full-width half maximum (FWHM) of sharp structure peaks along with no change in loss feature at 372.5 eV and the separation between these spin-orbit components is nearly 6 eV, indicating Ag present as a pure metal form in these PMCs. The people of these ages may be aware of the use of silver and about taking silver from their ore and making them coins. We could say that knowledge was transferred from one period to the next. Further, the overlapped binding energy curve of Cu for S-0 to S-VIII PMCs is shown in Fig. 5 (c). Here in the case of the copper, the binding energy curve for sharp distant doublets was observed at 933 eV for Cu²p_1/2 and 952.38 eV for Cu²p_3/2 in all the coins with separation in the spin-orbit coupling of about 19.40 eV with an intensity ratio of about 0.51. However, the Cu metal and Cu (I) oxide both have nearly the same binding energy i.e. 933 eV³⁸. The peak profile of Cu XPS spectra was not sharp enough as in the case of silver. The normalized curves for S-0 to S-VIII for Cu are shown in Fig. 5(d), indicating broadening in the peaks along with the appearance of satellite peak at 943 eV in some of the peaks indicating oxidation of surface copper and present as CuO, however, it is difficult to distinguish Cu with its oxidation state by XPS spectra. The peak intensity corresponding to the binding energy of lead was found small. The overlapped binding energy curve of Pb for different PMCs as shown in Fig. 5(e) shows peaks at 143.38 to 138.65 eV binding energy region with different intensities in which doublet structure corresponds to Pb⁴f_5/2 and Pb⁴f_3/2 energy states of Pb₂O₄ having spin-orbit separation was found at 4.73 eV. The normalized curve [Fig. 5(f)] is well overlapped on each other indicating Pb is present only in oxidized state.

Fig. 5.

Binding energy XPS curve for (a) Overlapped Ag3d, (b) Normalized Ag 3d, (c) Overlapped Cu2p (d) Normalized Cu2p, (e) overlapped Pb4f and (f) Normalized Pb4f curve.

In the case of S-0 Kashi PMC, an interesting observation was made towards the low binding energy side of XPS spectra. Well-separated doublet structure (3.7 eV separation) with remarkable intensity appears at 84.18 and 87.88 eV (supporting Fig. 1(c)). These binding energy peaks fall in the 4f region of XPS spectra of gold and are assigned to Au⁴f_7/2 (84.18 eV) and Au⁴f_5/2 (87.88 eV) components. As per ISO 15472:2010, the binding energy of pure gold is 83.95 eV. Similar binding energy for gold also appears in literature³⁸ and confirms the presence of gold in Coin 1 (S-0). Supporting Fig. 3(d) shows the close doublet binding energy curve with maxima at 161.38 and 162.48 eV corresponding to sulphur S²p_3/2 and S²p_1/2 respectively. The metal sulphide has a binding energy XPS peak at 161.5 eV [supporting figure S1(d) and S5 (d)]. There is a small possibility that some of the gold might be present as the-bound gold or Au₂S corresponding to BE.162.48 eV. The silver, copper, gold, sulphur and lead (as oxide) content is 67.3%, 5.3%,3.3%, 1.3% and 0.1% in PMC coin S-0. Further, Ag³d_5/2 peak exhibits very small chemical shifts, making it challenging to distinguish between different chemical states of Ag solely based on its binding energy curve in XPS measurements. To identify the chemical state of the metallic Ag or silver oxides (Ag₂O, AgO), and other silver compounds, the Auger parameter (α)^35–39 was calculated by combining the binding energy (BE) of the Ag ³d_5/2 peak with the kinetic energy (KE) of the Ag MNN Auger transition in different environments and compared the results with the standard reference for each PMC sample. The KE of Ag, Ag₂O and AgCl is 357.0 eV, 358.4 eV and 359.3 eV respectively³⁷. A comparison of the calculated and standard value of the Auger parameter for all nine PMC samples is shown in supporting Table 2. The Auger parameter of Ag (375.29 eV) found close to the reference value (375.30 eV)³⁹ of pure metallic Ag, indicates that surface Ag is not oxidised. Similar to Ag, Auger parameters were also calculated for different chemical states of Cu and Pb and correlated with their standard reported values as tabulated in supporting Tables 3 and 4 respectively. The kinetic energy of Cu, Cu₂O (Cu⁺¹) and CuO (Cu²⁺) are 918.0 eV, 917.5 eV and 916.5 eV respectively³⁷. The estimated Auger parameter value of 1850.78 eV was found close to the reference value³⁹ of pure metallic Cu value of 1850.60 eV. This shows that Copper is also present in its pure metallic form in these PMCs. Further, The kinetic energy of PbO (Pb²⁺) and PbO₂ (Pb⁴⁺) are 1024.5 eV, 1020.5 eV and 1017.0 eV respectively. The calculated Auger parameter of Pb is 1155.65 eV is close to the reference value 1155.70 eV, which is close to Pb₂O₃^36,37. The calculated Auger parameter for gold (with a binding energy of 84.17 eV) was found to be 2094.17 eV, which is close to the reference metallic state of Au ⁴f_7/2 (2094 eV). This suggests that the gold in the S-0 coin is in a purely metallic state. In contrast, the binding energy of 87.98 eV corresponds to sulfur-bound gold, with an Auger parameter of 2096.98 eV, approaching the reference value of 2092.5 eV. This indicates the presence of sulfur-bound gold on the coin’s surface. Additionally, for coin S-V, the Auger parameter for Al was measured at 1466.88 eV, which is close to the reference value for Al₂O₃ at 1464.50 eV.

On correlating the above experimental findings of PMCs, with archaeological data series, it gives valuable insights into ancient coins’ metallurgical properties, particularly those made from indigenous silver in northern India. This correlation is crucial for understanding the region’s economic history, as it reveals variations in silver content across different coin series. Silver, identified as the dominant element, decreased in percentage with increasing lead and copper in later series, highlighting advanced alloying techniques and metallurgical practices between the sixth and third centuries BCE. These coins’ varying Ag/Cu ratios suggest that they were made from cast metal pieces with specific compositions. The nine coins studied can be categorized into three groups based on their silver content: (i) 60% silver or more, (ii) 50% silver or more, and (iii) below 50% silver. Early coin series (S-0 to S-II) reflect a robust economy through their high silver content. However, political and administrative challenges led to a decline in economic conditions, evidenced by lower silver content in series S-III to S-V as shown in Fig. 4. According to P.L. Gupta and Hardaker²⁸, these coins were issued during the Shaishunaga dynasty up to the early phase of the Maurya dynasty, a period marked by political instability and Greek invasions. Series VI coins, issued during Emperor Ashoka’s reign, indicate a revitalized economy. The Arthashastra provides context for the value of Karshapana (PMC) in terms of fixed incomes¹⁵ for various social groups. Kautilya’s rules for mixing elements in silver Karshapana maintained high silver content in series VI coins until Ashoka died in 232 BCE. The subsequent lack of capable successors and territorial losses led to a rapid debasement of coinage post-Ashoka. Debasement resumed in later periods, with series VII and VIII coins showing very low silver content. Between 200 and 170 BCE, the expansion of the Indo-Greek Bactrian kingdom further impacted the Mauryan/Sunga economy in northwestern India. This likely led to the closure of Karshapana mints and the circulation of abundant Indo-Greek silver coinage. The invasion caused significant changes in coinage, including the debasement of certain Mauryan-style types, the cessation of silver PMCs, the emergence of local copies in base metals, and the rise of cast copper series. The presence of minor elements also indicates the ores contain these elements in the mines, from where they get raw material.

Conclusion

Finally, summarizing the preliminary results of XRD and XPS measurements of imperial PMCs (S-0 to S-VIII) of 600 to 200 BCE indicates that Ag is dominant along with some other elements considered impurities. XRD results of these PMCs reveal that metallic silver (Ag) found in the face-centred cubic (fcc) phase on the surfaces of the PMCs, along with a slight variation of nearly 0.05 Å in d-spacing and change in full-width at half maximum across samples S-0 to S-VIII. XPS measurement shows that Ag and Cu are present in pure metallic form, whereas the Pb on the surface of the coins gets oxidized and prevents the casting information of the chemical composition of PMCs. This shows a high level of metallurgical knowledge of alloying skills between the sixth and third centuries BCE. Pb increases the brittleness with concentration and may be used to determine the coins’ age. The carbon detected by XPS could originate from smelting, surface contamination, or environmental deposition. Also, the coins under consideration were issued at different times with technological changes. Intriguingly, Au has been exclusively discovered in Janapada PMC, contrasting with its absence in imperial PMCs. This pattern possibly underscores the connection between Janapada coins and the Vedic Nishka, which was essentially a gold bar. The declining silver content raised numerous questions regarding the economic climate, the increased demand for coins for exchange, and the maintenance of weight standards. Elemental analysis reveals that these coins have a non-uniform distribution of elements with varying percentages. It appears that the mines from where the raw materials were collected for minting the PMC are enriched with Ag, Cu, Pd and some other minor elements. However, the purpose of adding copper and lead may be either to lower the melting point or to meet the increased demand for coins in the Mauryan period and maintain the import of silver because of its price rise. In other words, to control the economy of the country. It is a valuable resource for archaeometallurgists when linked with established historical timelines and minting practices.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Amit Upadhyay: Data curation, Resources, Writing – original draft, Visualization, Neeraj Kumar Giri: Data curation, Formal analysis, Methodology, Validation, Rajiv Prakash: Formal analysis, MethodologyHirdyesh Mishra: Conceptualization, Investigation, Supervision, Resources, Validation, Writing – review & editing.

Data availability

Data is provided within the manuscript and supplementary information file.

Declarations

Competing interests

The authors declare no competing interests.