Waste Circuit Boards and Strategic Metals

01 / Analysis

What the 2020 study establishes

The authors characterize metals in waste circuit boards from phones, computers and televisions. They compare whole-board content, separated components and bare boards, then examine four particle-size fractions. The results support feedstock sorting and process investigation; they leave commercial recovery and delivered supply to be demonstrated.

Published in the September 2020 issue, volume 8(3), pages 590–602; accepted 16 October 2019. “Newer phones” means smartphones made after 2008 in this sample classification. These are historical laboratory measurements, with no reported industrial recovery trial or representative assay for present-day devices. [1]

• Literature comparison · printed 591
• Sampling and laboratory method · printed 592–594
• Whole boards and separated fractions · Tables 2–3 · printed 596
• Particle-size measurements · Tables 4–5 · printed 597–598
• Conclusions · printed 599–600

02 / Analysis

Identify the material before interpreting the number

Seven sample groups were selected from manually sorted waste boards in Croatia, averaging about 28 kg each. Categories distinguish older and newer phones, older and newer computers, CRT televisions and two LCD board functions. The paper does not identify a collection year or provide a device count for each group. [1]

Study workflow · printed 592–594 · procedures summarized from the original paper

Scroll horizontally to read all columns →

The paper does not provide numerical detection limits, replicate counts, certified-reference-material results or a complete analytical uncertainty budget. Values shown as < DL in size-fraction tables mean below detection, rather than proven absence. The apparatus and digestion description should not be treated as a validated commercial flowsheet. [1]

03 / Analysis

Compare the same tonne of board feed

The following tables reproduce the nine metal columns in Table 2. Precious metals use g/t; the six base metals use kg/t. These concentrations belong to circuit boards with components, rather than complete phones, computers or televisions. [1]

Table 2 · precious metals · g per metric tonne of circuit-board feed · printed 596 / PDF 7

Scroll horizontally to read all columns →

Per metric tonne of the sampled circuit boards, including their electronic components. Values are reported metal content, not metal recovered or content per tonne of complete devices. [1] Download dataset CSV

Table 2 · base metals · kg per metric tonne of circuit-board feed · printed 596 / PDF 7

Scroll horizontally to read all columns →

Per metric tonne of the sampled circuit boards, including their electronic components. Values are reported metal content, not metal recovered or content per tonne of complete devices. [1] Download dataset CSV

Table 1 is a separate literature compilation. Its silver mean of 0.1027% converts to 1,027 g/t, with a published range of 0–0.3800%, equivalent to 0–3,800 g/t. Those heterogeneous earlier observations are not the mean or range of the seven samples shown here. For conversions, 1% equals 10 kg/t or 10,000 g/t. [1]

Table 2 reports newer-phone gold at 1,071.35 g/t, while one sentence on printed page 595 gives 1,071.96 g/t. This page uses the table value; the weighted Table 4 calculation is within 0.01 g/t using the published rounded inputs. Published 0.00 palladium values are retained as table entries; the fraction tables include below-detection results. No zero should be interpreted as proof that every such board contains none. [1]

04 / Analysis

Follow both separated streams

The study separated components from computer and television boards, with both streams retained for assay. Phone boards could not be separated by this method. Component mass includes non-metals, so a component mass share is not a metal mass share. [1]

Table 3 · mass shares of the whole board (%) and precious-metal assays per tonne of each fraction (g/t)

Scroll horizontally to read all columns →

E = removed electronic components; B = remaining bare base plate. Each assay is per tonne of that fraction. Mass shares are percentages of the original board assembly. [1] Download dataset CSV

Table 3 · six base-metal assays · kg per tonne of each separated fraction

Scroll horizontally to read all columns →

E = removed electronic components; B = remaining bare base plate. Each assay is per tonne of that fraction. Mass shares are percentages of the original board assembly. [1] Download dataset CSV

For the signal-reception boards, silver is 215.06 g/t in components and 244.41 g/t in base plates. Weighting these assays by 64.97% and 35.03% of board mass gives about 225.34 g/t, consistent with Table 2. Counting both fraction assays as whole-board grades would overstate the contained metal. [1]

These concentrations describe where metals were measured. Demonstrating a separation benefit requires a material balance, treatment cost, dust and residue accounting, and the downstream recovery of both streams.

05 / Analysis

A rich fraction can hold a small share of the metal

These two Table 4 examples show why grade and metal mass need separate views. Older-computer components have their highest gold assay below 0.1 mm, while newer-phone boards have their highest gold assay in the 1–4 mm fraction. The phone sample is an exception to a simple “finer means richer” rule. [1]

Selected Table 4 inputs · mass share of each feedstock (%) and gold assay per tonne of that size fraction (g/t)

Scroll horizontally to read all columns →

C OLD(E) is separated older-computer components; CP NEW is integral newer-phone boards. Assays are per tonne of each size fraction. Contained-gold shares are calculated from published mass shares and assays, without a recovery factor. [1] Download dataset CSV

SilverWars calculation from Table 4 · share of contained gold in each selected feedstock (%) · not a recovery rate

Scroll horizontally to read all columns →

C OLD(E) is separated older-computer components; CP NEW is integral newer-phone boards. Assays are per tonne of each size fraction. Contained-gold shares are calculated from published mass shares and assays, without a recovery factor. [1] Download dataset CSV

For older-computer components, the < 0.1 mm fraction has a gold assay of 717.50 g/t but represents only 3.91% of component mass. Its calculated share of contained gold is about 14.22%. The larger 1–4 mm fraction contains about 57.51%. Both statements can be true without establishing a recovery yield.

Conclusions summarize tendencies across samples; the tabulated exceptions remain relevant. No consistent silver enrichment rule is established across all particle sizes. The calculated distributions use rounded published inputs and describe the selected feedstocks only. [1]

06 / Analysis

Account for the losses and the route to a buyer

1. Collect

   How much suitable board feed can actually be secured?

2. Separate

   Where does the metal go, including dust and residues?

3. Refine

   What metal-specific yield and product purity are demonstrated?

4. Deliver

   Who owns the output and can the buyer qualify it in time?

Evidence ladder · SilverWars analytical framework · the study supplies assay evidence at the first step

Scroll horizontally to read all columns →

Apply collection and board-separation factors only when starting with complete-device waste. A tonne of already separated boards has a different denominator. Do not multiply a circuit-board grade by the mass of all global e-waste, or apply a whole-stream recycling percentage as a silver recovery rate.

The strategic investigation follows collectors, processors, refining access and offtake rights across borders. A domestic collection program can still rely on overseas separation or refining. Evaluate the economics of the full metal mix alongside environmental controls and accountable residues.

07 / Analysis

Secondary resources compete through industrial systems

Examine silver data and availability definitions →
Explore military and aerospace silver specifications →
Read the analysis of U.S. supply alternatives →

08 / Analysis

Turn the assay into a testable investigation

1. What feedstock can be secured?

   Which board types, equipment generations and annual quantities are covered by collection or purchase contracts?

   Evidence to seek Collection records, device mix, sorted-board weights, competing reuse routes and contract rights.

2. How representative is the grade?

   Do independent, repeated batch assays reproduce the reported values for current feedstock?

   Evidence to seek Sampling plans, accredited laboratory reports, detection limits, reference materials and uncertainty.

3. Where does each metal go?

   Do the mass balances account for both component and base-plate streams, fines and treatment residues?

   Evidence to seek Fraction weights, assays and metal-specific balances across actual operating campaigns.

4. Who controls refining and delivery?

   Can the processor deliver a qualified product through a permitted route within the buyer’s required time?

   Evidence to seek Facility ownership, process capability, shipment consents, settlement terms and customer qualification.

5. Does the full process sustain itself?

   Do payable metals cover collection, separation, refining, transport and residue treatment under documented conditions?

   Evidence to seek Realized settlements, operating costs, contaminant penalties, energy use and environmental performance.

09 / Analysis

Trace every measurement and interpretation

Tables and charts are SilverWars adaptations of data reported by Anić-Vučinić, Bedeković, Šarc and Premur in JSDEWES 8(3), 590–602 (2020), DOI 10.13044/j.sdewes.d7.0312. Values from Tables 2–4 retain their sample basis. Figure layouts and contained-gold calculations are new; no author or publisher endorsement is implied.

Publisher’s CC BY 3.0 attribution policy ↗
Creative Commons Attribution 3.0 licence ↗

A strategic resource intelligence assessment reviewed on 3 October 2026. Original measurements describe seven Croatian waste-PCB sample groups and selected separated fractions. Contemporary evidence retains its measurement years and policy scope. This record does not establish a global board grade, a commercial recovery yield, a current shipment approval or available refinery inventory.

Reported facts retain their source dates and definitions. The resource implications and research questions are SilverWars analysis. Policy announcements and critical-mineral designations do not by themselves prove adequate inventory or an inevitable shortage.

1. 01

   Original peer-reviewed laboratory study

   Determining Metal Content in Waste Printed Circuit Boards and their Electronic Components

   Anić-Vučinić, Bedeković, Šarc and Premur · JSDEWES / SDEWES Centre · September 2020

   Evidence and source limits

   Locator

   Printed 590–602 / PDF 1–13; Table 1 literature comparison p.591; method pp.592–594; Tables 2–3 p.596; Tables 4–5 pp.597–598; conclusions pp.599–600.

   Supports

   Historical board and fraction assays, sample definitions, laboratory method and observed exceptions. Original table values support the new chart layouts and explicitly calculated contained-gold distributions.

   Limit

   Seven selected Croatian sample groups; collection date and device counts are not established. No complete uncertainty budget or commercial recovery trial. Table 2 and one narrative gold value differ; fraction-table < DL entries require detection-limit interpretation.

   Reviewed

   October 3, 2026

2. 02

   Primary institutional global statistics

   Global E-waste Monitor 2024 · release and principal findings

   UNITAR and ITU · March 20, 2024

   Evidence and source limits

   Locator

   Release of 20 March 2024; 2022 generation and documented formal collection/recycling figures.

   Supports

   62 million tonnes of e-waste generated and 22.3% documented formally collected and recycled in 2022.

   Limit

   All e-waste, not only PCBs. Report year differs from measurement year. Formal collection/recycling mass is not recovered silver mass or a current 2026 flow.

   Reviewed

   October 3, 2026

3. 03

   Primary international treaty guidance

   E-waste Amendments · Secretariat questions and answers

   Basel Convention Secretariat · Date not established · Effective January 1, 2025

   Evidence and source limits

   Locator

   Questions 1–5: scope, entries A1181 and Y49, entry into force and non-acceptance; questions 6–12: definitions and controls.

   Supports

   Prior-informed-consent scope for e-waste movements, component and processing-residue coverage, and exceptions to amendment acceptance.

   Limit

   The FAQ has no established publication date. National definitions, Party status, restrictions and shipment-specific approvals still require verification.

   Reviewed

   October 3, 2026

4. 04

   Primary legislation

   Regulation (EU) 2024/1252 · Critical Raw Materials Act

   European Parliament and Council · Official Journal of the European Union · May 3, 2024 · Effective May 23, 2024

   Evidence and source limits

   Locator

   Article 2 definitions; Article 5(1)(a)(iii), 2030 recycling-capacity benchmark; Annex I material scope; Article 49 entry into force.

   Supports

   The 2030 Union recycling-capacity benchmark relative to annual strategic-raw-material consumption.

   Limit

   Capacity benchmark, not achieved output or a 25% PCB/silver recovery requirement. This analysis does not infer completed implementation or a permit for any project.

   Reviewed

   October 3, 2026

5. 05

   Primary government statistics

   Mineral Commodity Summaries 2026 · version 1.3

   U.S. Geological Survey · February 2026 · Edition updated May 2026

   Evidence and source limits

   Locator

   Silver printed 172–173 / PDF 176–177: 2025 secondary production, recycling share, net import reliance and addition to the 2025 list.

   Supports

   2025 U.S. silver recycling and consumption estimates, import reliance and silver’s U.S. critical-mineral designation.

   Limit

   All new and old silver scrap, not PCB-only recycling. Consumption share is not process yield, and annual statistics do not establish deliverable inventory.

   Reviewed

   October 3, 2026