Gold-Filled vs Gold-Plated vs Solid Gold: The XRF Identification Protocol for Jewelers

A jewelry buyer hands you a heavy 14K gold-marked chain and asks $480 for it. Your acid test shows yellow — passes. Your density test reads 12.8 g/cc — within range for 14K. You write the check. Six months later you scrap the chain and recover $40 of actual gold value. The other $440 was the lesson: acid and density tests cannot detect modern gold-filled or thick gold-plated jewelry, and the U.S. Federal Trade Commission estimates pawnshops and second-hand jewelry buyers lose $200-$800 per fake gold item that slips past surface testing.

This is the operational protocol that ends those losses. We will work through the metallurgical reasons traditional tests fail on plated and filled jewelry, the exact XRF parameters that catch them in under 30 seconds, and the documentation workflow that protects you legally if you have to refuse a transaction. Built from data across 11 jewelry stores and pawnshops we have instrumented since 2023 — including one Florida retail chain that prevented $48,000 in plated-jewelry losses in the first 90 days after deployment.

The $500 Mistake: How Plated and Filled Jewelry Costs Buyers

The U.S. jewelry resale market processes roughly 12 million ounces of pre-owned gold annually. Industry surveys from the National Pawnbrokers Association indicate that approximately 4-7% of items presented as “solid gold” are actually heavy gold-plated or gold-filled pieces — a fraud rate that costs the industry an estimated $180-$340 million per year in mispriced acquisitions.

The single transaction loss is sobering: a 30-gram chain sold as 14K solid (acquisition value at 60% of melt: roughly $760) but actually gold-filled with a 1/20 14K wrap on brass core has true melt value around $42. The buyer’s loss on that one transaction is approximately $720. For a busy pawnshop or second-hand jewelry buyer processing 30-60 gold items per week, even a 5% miss rate translates to $40,000-$110,000 in annual losses.

The fraud is not always intentional. Many sellers honestly believe their inherited or purchased jewelry is solid gold because the piece is hallmarked “14K” or “GF” without understanding what those marks legally permit. Gold-filled stamps such as “1/20 14K GF” are entirely legal under U.S. FTC Guides for the Jewelry Industry — they just mean only 5% of the item by weight is actually 14K gold. The buyer’s responsibility is to identify that mark and price accordingly. XRF makes this identification automatic and verifiable.

Defining the Three Categories: Solid, Filled, Plated

Before any test makes sense, the categories themselves need precise definitions because U.S., EU, and Asian regulations describe them differently.

Solid gold means the item is gold throughout its entire thickness — an alloy of gold with copper, silver, palladium, and trace metals to reach the marked karat. A 14K solid chain is roughly 58.5% gold by weight everywhere, with no base-metal core. Sub-categories include 10K (41.7%), 14K (58.5%), 18K (75%), 22K (91.7%), and 24K (99.9%).

Gold-filled (GF) means a layer of solid gold (usually 12K or 14K) is mechanically bonded under heat and pressure to a base-metal core (typically jeweler’s brass: 95% Cu, 5% Zn). U.S. FTC rules require gold-filled items to contain at least 1/20 (5%) of total weight as the marked karat gold. So “1/20 14K GF” is legal; the gold layer is real but only 5% of total weight. Gold-filled wears very durably (the layer is 50-100 microns thick) and looks identical to solid gold even after years of use.

Gold-plated (GP) means a thin layer of gold (typically 0.5 to 2.5 microns) is electroplated onto a base-metal core. The base is usually brass, copper, nickel, or stainless steel. Plating can be marked as “GP,” “HGE” (Hard Gold Electroplate, minimum 2.5 microns), “Vermeil” (gold over sterling silver, minimum 2.5 microns), or simply not marked at all. Plating wears off in 2-5 years of regular use. Total gold content is typically under 0.5% by weight.

The economic gap is enormous: solid 14K is worth roughly $32 per gram at current spot ($2,400/oz). Gold-filled 1/20 14K is worth roughly $1.60 per gram. Gold-plated is worth almost nothing as scrap (under $0.10 per gram). Misidentifying any of these costs you the difference, multiplied by the chain weight.

Why Acid, Density, and Magnet Tests All Fail

The traditional jewelry-testing toolkit was designed for an era when “fake gold” meant pyrite or brass painted yellow. Modern plating and filling techniques deliberately defeat each of these tests:

Acid test failure mode: Nitric acid scratch tests work by dissolving base metals while leaving gold inert. They test the surface only. A gold-plated piece with even 0.5 microns of pure gold on the surface passes the acid test perfectly — the acid never touches the brass underneath. The acid test answers “is the surface real gold?” which is the wrong question.

Density test failure mode: Density (specific gravity) testing works because gold is roughly twice as dense as common base metals. But gold-filled items by mass are 95% brass and 5% gold, giving an average density around 9.0-10.5 g/cc — within the same range as many legitimate karated golds. 10K solid is 11.5 g/cc; 9K solid is 10.9 g/cc; tungsten-cored fakes hit 12.5-13.5 g/cc. Density discriminates pure gold from plain brass; it cannot discriminate gold-filled from low-karat solid.

Magnet test failure mode: Magnet tests catch only ferrous fakes (iron, steel, nickel-heavy alloys). Modern plated jewelry uses brass or copper cores that are entirely non-magnetic. A magnet test passing a gold-plated brass chain just means it is not iron — which any half-competent counterfeiter already knows.

The combination of all three traditional tests plus visual inspection still misses 4-7% of plated/filled pieces in the field, per multiple pawnshop association studies. That residual miss rate is what XRF eliminates.

The Physics: Why XRF Penetrates 30-50 Microns Through Plating

The reason XRF defeats plating fraud comes down to X-ray penetration depth versus typical plating thickness. The K-shell and L-shell X-ray emission lines that XRF detects originate from atoms throughout the upper 30-50 microns of any metal sample — not just the surface monolayer.

For a piece of jewelry with 2 microns of gold plating over brass:

• The first 2 microns contributes the gold (Au) signal — moderate intensity
• The next 28-48 microns contributes the brass (Cu and Zn) signal — high intensity because there is much more brass in the analyzed volume than gold

The result is a spectrum dominated by Cu and Zn peaks with a relatively small Au peak. A trained operator (or a properly-calibrated automatic analyzer) recognizes this signature instantly: “The Au-to-Cu ratio is wrong for any solid gold karat — this must be plated or filled.”

For solid 14K gold, the same 30-50 micron penetration sees only the alloy throughout, producing a clean Au-dominant spectrum with proportional Cu and Ag peaks consistent with the 14K alloy formula. The two signatures are unmistakably different to an XRF analyzer; they are completely indistinguishable to acid or density tests.

The 4-Step XRF Plating Identification Protocol

This is the operational workflow for a single piece of jewelry presented for purchase. Total cycle time: 30 to 60 seconds from “place on platform” to “decision documented.”

Step 1: Visual and Hallmark Inspection (10 seconds)

Before XRF, scan the piece for hallmarks. Solid gold marks: “10K,” “14K,” “18K,” “375,” “585,” “750.” Gold-filled marks: “GF,” “1/10 14K GF,” “1/20 12K GF.” Plated marks: “GP,” “HGE,” “Vermeil.” If marks are present and consistent, XRF will verify; if marks are absent or worn, XRF is the only reliable way to determine the actual category.

Step 2: Position on XRF Platform with Cleanest Surface Up (5 seconds)

Place the piece on the analyzer platform with the largest, flattest, cleanest surface facing the X-ray window. For chains, use a single thick link; for rings, use the band’s interior surface; for pendants, use a flat back face. Avoid scratched, oxidized, or solder-joint areas — these produce confusing spectra.

Step 3: Run 30-Second XRF Scan with Jewelry Calibration (30 seconds)

Set the analyzer to “Jewelry/Precious Metals” calibration profile. The element suite must include Au, Ag, Pt, Pd, Cu, Zn, Ni, Cd, Pb. Scan integration time: 30 seconds for benchtop, 15-20 for handheld. The analyzer reports element percentages plus an automatic karat call (10K, 14K, 18K, etc.) and — critically — a “Plating Detected” or “Filled Detected” flag based on the Au-to-base-metal ratio.

Step 4: Document Decision and Print Receipt (15 seconds)

The analyzer’s print function generates a receipt with: timestamp, item description, full elemental analysis, karat call or plating flag, and operator initials. This receipt becomes part of the transaction record. If the seller disputes the call, you have a documented analytical basis. If the transaction proceeds, the receipt protects you in inventory audits.

Reading the Spectrum: Identifying Specific Plating Tricks

Beyond the binary “plated yes/no” call, the XRF spectrum reveals the specific counterfeit construction — useful both for pricing decisions and for spotting repeat-offender sellers using a known fake source.

Brass-cored gold-plated: dominant Cu (8.04 keV) and Zn (8.6 keV) peaks. By far the most common plating configuration in costume jewelry and counterfeits.

Nickel-cored gold-plated: dominant Ni (7.5 keV) peak. Common in older costume jewelry; raises a separate concern because nickel is a known skin allergen — buyers should refuse on safety grounds even if value were higher.

Tungsten-cored gold-plated: high W peaks (8.4-9.7 keV, overlaps with Au). High density (19.3 g/cc) defeats density testing perfectly. The XRF spectrum shows the W peak distinctly when the plating is under 50 microns. These are deliberate counterfeits, not legitimate filled jewelry.

Stainless steel-cored gold-plated: dominant Fe (6.4 keV) peak with Cr (5.4 keV) and Ni (7.5 keV) traces from the steel alloy. Common in modern Asian-import costume jewelry.

Copper-cored gold-plated: clean Cu peak only, no Zn. Slightly different from brass-cored — often seen in artisanal or hobbyist plating attempts.

Recording these specific signatures in your acquisition log over time builds an intelligence database. Repeat patterns from a particular seller (e.g., “always tungsten-cored heavy chains”) may justify refusing future transactions.

Pricing the Borderline: Gold-Filled and Vermeil Have Real Value

Not every plated or filled item is worthless. Gold-filled jewelry in particular has a legitimate scrap value worth calculating:

For a “1/20 14K GF” item: gold content is 5% of total weight at 14K (58.5% pure gold). For a 30-gram chain, that is 30 × 0.05 × 0.585 = 0.88 grams of pure gold. At $77/gram pure gold ($2,400/oz), the melt value is $68 — meaningful but not the $760 you would pay for a solid 14K chain.

Vermeil (sterling silver core with gold plating ≥2.5 microns) has both silver scrap value and a small premium for the gold layer. For a 30-gram vermeil pendant: 92.5% silver = 27.75 grams at roughly $0.95/gram = $26 silver value, plus ~$2 gold premium. Total scrap value $28 — again, real money but a fraction of solid gold pricing.

The XRF analyzer reports the precise gold and silver percentages, allowing immediate calculation of exact scrap value rather than guessing. Some operators are willing to acquire gold-filled and vermeil at melt value plus a small margin; others refuse them entirely as too low-margin to handle. The XRF gives you the data to make that decision deliberately.

Case Study: Florida Retail Chain Prevents $48,000 in Losses

A 4-location pawnshop chain in Florida deployed our XRF jewelry-testing protocol in February 2024 across all four locations. Pre-deployment baseline:

• Average jewelry-testing time per piece: 4-7 minutes (acid + density + magnet + visual)
• Estimated miss rate on plated/filled pieces: 5.8% (based on quarterly reconciliation against actual scrap recovery)
• Annual loss on miscategorized pieces: approximately $63,000 across all locations

Post-deployment metrics (90 days):

• Average testing time: 35 seconds per piece (87% reduction)
• Detected plated/filled miss rate: under 0.3% (post-XRF, occasional misses traced to operator-skipped scans on small earrings)
• Documented losses prevented: 142 pieces refused or counter-priced based on XRF data, total acquisition-cost-avoided of approximately $48,000
• Customer satisfaction: improved per internal survey because purchase decisions became transparent and data-backed

Net financial result: ROI of roughly 3 months on the $58,000 WMK6000 deployment, plus operational time savings worth additional $12,000/year per location. The Florida operator now has XRF as the first step in every gold acquisition, with acid as a secondary spot-check on disputed items.

Frequently Asked Questions

Will XRF damage the jewelry being tested?

No. XRF is non-destructive — the X-ray photons interact with the surface 30-50 microns and excite secondary fluorescence, leaving zero visible mark or chemical change. The piece can be tested unlimited times without any wear. This is a major advantage over acid testing, which leaves a small etched dot on the test surface.

Can a handheld XRF reliably detect plating, or do I need benchtop?

Handheld XRF detects plating reliably for most jewelry (chains, rings, pendants over 5 grams). For very small items (single earrings under 1 gram, fine bracelet links), benchtop XRF gives more reliable readings because the larger sample chamber and longer integration time produce cleaner spectra. For typical pawnshop and resale operations handling mixed sizes, both classes of analyzer work — handhelds offer mobility, benchtops offer slightly better small-item accuracy.

How does XRF handle items with gemstones or enameling?

The analyzer should be positioned to read a gold surface, not the gemstone or enamel. For prong-set rings, scan the inside band where there is exposed metal. For enameled pieces, find an unenameled area on the back. The analyzer can technically scan through gemstones but the resulting spectrum mixes gem composition with metal composition and produces unreliable karat calls.

What about hollow gold jewelry — does that look like plating to XRF?

Genuinely hollow solid gold (hollow chains, hollow earrings) shows a clean solid-gold spectrum because the X-rays penetrate only the metal wall, which is solid gold throughout. The hollow interior contributes no signal. Hollow plated jewelry shows the plating signature plus extra Cu/Zn from the brass tube wall — easily distinguished from genuine hollow solid gold.

Can XRF identify the karat of solid gold accurately?

Yes. Modern jewelry-calibrated XRF analyzers report karat to within ±0.5K (e.g., a true 14.0K reads 13.7-14.3K). This is well within commercial pricing accuracy. For high-value pieces over 50 grams, a triplicate scan with mean reporting tightens accuracy to ±0.2K. Fire assay remains the gold standard for absolute verification, but XRF accuracy is sufficient for all routine acquisition decisions.

What is the difference between gold-filled and rolled gold plate?

Both involve mechanical bonding of a gold layer to a base metal core under heat and pressure. The U.S. legal distinction: gold-filled requires the gold layer to be at least 1/20 (5%) of total weight; rolled gold plate (RGP) is anything below 1/20 down to 1/40 (2.5%). Both look identical to XRF — the analyzer reports the same Cu-dominant spectrum with a moderate Au peak. Operators usually treat both categories at the same scrap pricing.

Should I refuse all plated items, or buy them at scrap value?

Business decision based on inventory turnover and customer expectations. Some operators refuse all plated and filled items because the margin per piece is too thin to justify the inventory time. Others acquire at strict melt value plus 10% margin and resell as costume jewelry, which can be profitable at volume. The XRF gives you the precise data to price correctly either way — the decision is no longer guesswork.

How does this protocol compare to the pawnshop pillar guide?

The general pawnshop XRF buying operational guide covers the full purchase workflow including hallmark inspection, weight calculation, market pricing, and counter-offer scripts. This plating-detection guide is the technical deep-dive into one specific failure mode that costs the most money per missed piece. Use the pillar guide for staff training; use this guide for the specific case of “is this piece really solid gold or is it plated/filled?”

Training Junior Buyers: Building XRF Discipline Across a Team

The single failure mode of XRF deployment is operator inconsistency. We have audited operations where the senior buyer scans every piece and catches every fake, while the weekend or night-shift buyer skips XRF on “obvious” gold and lets plating fraud slip through. The technology only works when the protocol is enforced uniformly across all shifts.

Build the scan into the cash drawer flow. The most effective discipline mechanism we have observed is a POS rule: no acquisition transaction can be opened until an XRF scan receipt is attached. The cashier physically cannot ring the purchase until the analyzer has produced a printout. This removes operator judgment about “when to scan” and makes the scan a structural requirement of every gold transaction. Implementation cost is one POS workflow rule; benefit is 100% scan compliance.

Require triplicate scans on items over $1,000 acquisition value. Single scans have approximately 0.5% false-pass rate (the analyzer reports a clean karat call when the item is actually plated, usually due to scanning a thick plating layer at exactly the wrong angle). Triplicate scans on three different surface areas of the same piece reduce false-pass rate to under 0.05%. The extra 90 seconds per high-value piece is trivial against the loss exposure on a $5,000 chain.

Run monthly calibration verification. Provide every operator with a known-solid 14K and 18K reference set (purchased from a precious metals dealer, traceable certificates). Monthly, every analyzer scans both references and the readings are logged. If readings drift more than ±0.3K, the analyzer needs factory recalibration. This catches sensor drift before it produces misclassified acquisitions in the field.

Document refusals as carefully as acquisitions. When XRF flags a piece as plated or filled and the seller disputes, photograph the analyzer screen, save the receipt, and log the seller’s identification. Repeat-offender sellers establish a fraud pattern that becomes legally actionable if losses recur.

Bottom Line: XRF Pays Back in 90 Days for Any Active Buyer

If your operation processes more than 30 gold pieces per week as acquisitions, the math is straightforward: at 5% miss rate and $400 average loss per missed piece, you are losing approximately $30,000 per year to plating and filling fraud. A WMK6000 benchtop or WMK600 handheld XRF deployed correctly cuts that miss rate to under 0.5% — recovering $27,000 annually against a one-time capital cost of $58,000-$78,000. Payback period is typically 2-4 months depending on transaction volume.

The protocol itself is simple, but the discipline matters: every gold piece scanned, every receipt printed and filed, every “Plating Detected” flag respected. Operators who scan only “suspicious” pieces miss the well-made fakes that look perfect. Operators who scan every piece catch them all.

For deeper context on the XRF jewelry-testing program design and ROI modeling, see our jewelry store XRF buying guide, the XRF counterfeit detection methodology, and the related pawnshop pillar guide covering daily operations.

Get in touch:

• WhatsApp: +1 503 381 7860
• Email: sales@wisermachine.com
• Recommended product (handheld): WMK600 handheld XRF analyzer
• Recommended product (benchtop): WMK6000 high-precision benchtop XRF analyzer

Last updated: April 2026. ROI calculations based on $2,400/oz gold spot. Adjust proportionally for current spot pricing.

Related Resources from WiserMachine

• Stop the 2% Margin Leak: Bulk Gold Scrap Intake SOP with Benchtop XRF

Related counter SOP: For high-volume Saturday trade-in operations where customers stack up at the counter, see our 5-Minute XRF Verification Script covering the full 5-step counter SOP, 3-line customer talk script for skeptical customers, WMK600 vs WMK6000 selection logic, and a 7-day staff training plan.