Views: 0 Author: Peter Cui Publish Time: 2026-06-04 Origin: Mitour Silicone
Table of Contents
TL;DR — This is not a product catalog or a capabilities brochure. It is an unfiltered account of what we have learned running a silicone manufacturing operation in Shenzhen for 21 years — from our founding in 2005 through the Walmart audit in 2019, the Disney licensing milestone in 2022, and the development of our Solid-to-Liquid Refinement Technology. As our CEO often reminds the team: the factory is a teaching machine if you are willing to learn from it every single day. This article compiles those lessons for the buyers, brand founders, and procurement directors who are deciding whether to trust us — or any silicone manufacturer — with their brand.
Mitour Silicone was founded in 2005 and has operated in Shenzhen for 21 years.
We operate a 4,500 m² facility with 100+ employees, 20+ LSR injection machines, and 30,000 units/day capacity.
We have shipped to 50+ countries; Walmart and Target became clients in 2019; Disney licensed us in 2022.
Our 300+ patent portfolio includes CN113650314A (antimicrobial silicone) and CN114015239A (infant-grade silicone refinement).
The single most expensive mistake in silicone manufacturing is choosing a factory based on price instead of process control.
The three things that separate a 21-year factory from a 3-year factory: compound traceability, mold maintenance discipline, and the willingness to say no to an order you cannot execute well.
If you read nothing else on this site, read this article.
As our CEO often reminds the team, we did not start in 2005 with a vision statement and a venture capital round. We started with a concrete floor, four used HTV compression presses, and a phone with a SIM card that still felt like a novelty.
Shenzhen in 2005 was already the manufacturing capital of the world — or at least the part of the world that made things that ended up on shelves. The Pearl River Delta corridor from Shenzhen north through Dongguan to Guangzhou was running at full industrial adolescence: factories opening faster than infrastructure could keep up, labor arriving from inland provinces faster than housing could be built, and export orders flowing faster than anyone had capacity to fill.
Silicone was a niche then. Most buyers in Europe and North America associated "silicone" with adhesive sealant or medical implants. Consumer silicone — kitchenware, baby products, promotional goods — was just beginning its trajectory from novelty to mainstream. We saw that coming. Not through sophisticated market analysis but through the simple observation that every time a product was made of silicone instead of plastic, people liked it better. It did not break. It did not harbor bacteria the way plastic did. It survived a drop, a dishwasher, a child, a decade.
The first years were operationally unglamorous. We bought silicone compound from a domestic supplier whose quality was adequate but documentation was sparse. We cut molds in medium-grade steel that wore faster than we wanted. We had a quality control process that was, in retrospect, more optimism than engineering.
Two things saved us from the mediocrity we might have settled into. First, we found an early buyer — a European kitchenware importer — who refused to accept our first shipment because our LFGB migration test was 8% above the limit on acetic acid simulant. He did not threaten legal action. He sat with us and explained, for three hours, why the limit existed and what it meant for his customers. We retested. We passed. We changed our compound supplier. That European importer is still a client.
Second, we decided early that mold ownership was going to be a principle, not just a capability. We invested in our own CNC equipment in 2007 rather than sending tooling to a third-party mold maker. That decision — expensive and counterintuitive when we were cash-thin — is probably the single most consequential choice we made in the company's first decade. In-house tooling means in-house accountability. It means when a mold develops flash at 15,000 shots, we repair it ourselves at 8:00 the next morning. We do not wait for a subcontractor's schedule.
From 2005 to roughly 2014, our material traceability system was — let us be honest — a spreadsheet. A well-maintained spreadsheet kept by a conscientious warehouse manager, but a spreadsheet nonetheless. We knew which compound went into which production run. We could trace a batch if pressed.
What we could not do, in 2012, when a buyer's EU customs lab failed a shipment of kitchen silicone on a LFGB §30/31 volatile siloxane test, was identify within four hours exactly which compound lot was the source of the problem. It took us 36 hours. By that time, the buyer had missed a container booking window, the shipment was in customs hold, and the relationship — which had taken three years to build — was damaged in ways that took another two years to repair.
As our CEO often reminds the team when training new QC staff: traceability is not bureaucracy. It is the fastest way to solve a problem before it becomes a catastrophe.
Today, our lot numbers travel through every document in the chain: incoming CoA, production batch record, QC inspection log, shipping documentation. If you call us today and tell us a product from a 2023 shipment has failed a migration test, we can tell you the compound supplier, the lot number, the test report that compound was validated against, the machine it ran on, the operator shift, and the QC inspector who signed the outgoing batch. In under 20 minutes.
That capability was not created overnight. It required investment in ERP software, process redesign, and cultural change — convincing production operators that filling out a digital batch record was not paperwork bureaucracy but the thing that protected their jobs and the company's contracts.
The European kitchenware buyer who refused our first shipment in 2005 taught us one lesson. The 2012 customs hold reinforced it the expensive way. Both were necessary.
In the spring of 2019, a team from a Walmart-approved audit body arrived at our Building 2 LSR injection floor with 147 control points on a clipboard and a 48-hour window to work through them. The audit covered SMETA 4-Pillar social compliance, technical manufacturing capability, and product compliance documentation simultaneously.
As our CEO often reminds the team when this story comes up: the Walmart audit was not the moment we became a quality manufacturer. We had been working toward this standard for years. The audit was the moment someone measured it with precision and gave us a number.
We passed. One non-conformity: a color mixing log that did not cross-reference compound lot numbers. Minor. The corrective action took 14 days to implement and verify. Since then, not a single Walmart audit has issued a non-conformity to our system.
But the more important outcome was less visible. Going through a 147-point audit in 48 hours under real scrutiny — not self-assessment, but independent verification — taught us what we had and what we did not have. The things we thought were documented well often were documented adequately. The things we were most confident about occasionally had a gap the auditor's methodology exposed.
We came out of that audit with a corrective action list that included items we had not thought to look at. We worked through every item even though only one was formally required. That is what a quality management system should do: make you better, not just compliant.
Walmart becoming a client in 2019 opened doors — with Target in the same year, and with other major retailers who take Walmart approval as a reliable proxy for supplier capability. But the real value of the Walmart relationship is not the purchase order volume. It is the discipline it demands. Every year, there is another audit, another review cycle, another set of questions to answer with data. That ongoing scrutiny makes us better at what we do.
Becoming a licensed Disney manufacturer — which we achieved in 2022 — is not a credential you buy or negotiate into. It is a credential you earn through a qualification process that is among the most thorough in consumer goods manufacturing.
Disney's licensing program evaluates factories on: production capability (can you make what we need?), quality systems (will it consistently meet spec?), social compliance (SMETA 4-Pillar audit required), product safety documentation (EN 71 for toys; LFGB or FDA for applicable categories), and intellectual property protection (secure design files, controlled access to licensed artwork, NDA compliance).
The IP protection element is specific to Disney: their licensed artwork files are commercially extraordinarily sensitive. Disney requires documented controls over who can access digital design files, where they are stored, how they are transmitted, and how they are destroyed after a licensing agreement ends. We implemented a document control system that satisfied Disney's requirements in Q3 2021, went through a capability assessment in Q4 2021, and received manufacturing license in Q1 2022.
As our CEO often reminds the team: the Disney certification was not a reward for past performance. It was an invitation to a higher standard of performance going forward. Every shipment of licensed Disney products is reviewed against Disney's quality assurance protocol. Every season, there is a new product line, new specifications, new compliance requirements. Being a Disney manufacturer is not a status — it is an ongoing commitment.
The practical impact for buyers who are not Disney brand holders: the systems, documentation disciplines, and quality controls we built to meet Disney's standards operate across our entire production floor, not just on Disney-licensed orders. When you order from Mitour Silicone, you are operating within a quality system designed to satisfy the most demanding entertainment licensor in the world.
In conventional silicone manufacturing, starting from solid HTV gum compound to a finished product requires multiple handling steps: two-roll milling to compound the base polymer with fillers and crosslinker, sheeting, preform cutting, mold loading, press curing, post-curing. Each step introduces handling, each handling step introduces contamination risk, and each contamination risk is a potential product safety issue for applications where purity is paramount.
Infant-grade silicone products are the most demanding application in our portfolio. A baby bottle nipple made of silicone is in contact with expressed breast milk or formula, in contact with an infant's mouth for hours per day. Any extractable volatile from an inconsistently processed compound is not an abstract regulatory concern — it is a real risk to a real infant.
Around 2017, our technical development team began working on what eventually became Solid-to-Liquid Refinement Technology (a proprietary process developed in-house by Mitour Silicone). The core concept: develop a processing pathway that takes a carefully selected solid silicone raw material and refines it through a controlled conversion process to a liquid-phase compound with LSR-equivalent purity and processability, without the material overhead of conventional LSR's two-part metering system.
The development process ran for approximately four years, produced multiple patent filings, and ultimately resulted in patent CN114015239A — our infant-grade silicone refinement process patent. This patent protects the core conversion methodology.
What Solid-to-Liquid Refinement Technology delivers in practice:
Reduced volatile content: the refinement process drives off residual cyclics (D4, D5 cyclosiloxanes) to levels below 0.5 ppm, significantly below the thresholds that trigger LFGB §30/31 failures.
Improved platinum catalyst distribution: the conversion step homogenizes catalyst distribution in the compound more uniformly than standard LSR metering achieves in rapid-mix conditions.
Traceability: because the solid raw material input is more easily lot-coded and tracked than liquid LSR drums, material traceability at the compound level is simplified.
Application scope: products processed via this technology carry an additional layer of documentation substantiating their suitability for infant-contact applications, which we use in our baby product technical files for EN 71 compliance and FDA 21 CFR 177.2600 substantiation.
This technology is not a marketing claim. It is a patented process (CN114015239A) that sits in our production documentation and is referenced in our QC protocols for any product category where we invoke it. Buyers who want to understand the technical details can request our process overview document.
As our CEO often reminds the team: the most expensive word in manufacturing is "yes" said to the wrong order.
The wrong material, the wrong client. In 2011, a buyer asked us to produce silicone-coated fabric for an outdoor furniture application. We said yes because the revenue was attractive and because we convinced ourselves that silicone coating was close enough to silicone molding that we could figure it out. We could not. The coating adhesion failed in accelerated UV testing. The order was reworked twice, shipped late, and delivered at break-even after accounting for the rework cost. We do not do silicone-coated fabric. We should have said no in 2011.
The wrong volume, the wrong timeline. In 2016, a US promotional goods buyer asked us to produce 200,000 silicone wristbands in 10 days. Normal lead time for this quantity: 20–25 days. We said yes — we had the capacity, theoretically — and spent two weeks running double shifts on HTV presses, compromising our regular customers' orders to prioritize the rush. Quality was acceptable. Customer satisfaction was not — they found flash levels at the higher end of our tolerance, which was the inevitable result of running presses at their limit without the scheduled maintenance windows. We now quote rush surcharges that are intentionally high enough to make over-commitment less appealing.
The right product, the wrong compound. In 2018, a buyer asked us to use their nominated compound supplier for a food-contact kitchen product. We tested the compound; migration values were marginal — technically within LFGB limits, but at 92% of the limit on one simulant. We flagged this to the buyer; they pushed us to proceed on the grounds that "92% is still compliant." We proceeded. Six months later, a German retailer's incoming quality check ran an independent migration test and failed the batch — a different test protocol, slightly different temperature, the 8% safety margin was not enough. We absorbed half the cost of the failed shipment because our engineers had flagged the risk and we had run the order anyway. We now decline to use nominated compound suppliers for food-contact applications without running our own full LFGB panel first. Non-negotiable.
Three expensive lessons. All avoidable. All valuable.
As of 2026, Mitour Silicone holds 300+ granted patents, including 3 core invention patents. These are real, registered patents — not a marketing number inflated by counting utility model registrations alongside invention patents. The three core invention patents include CN113650314A (antimicrobial silicone composition) and CN114015239A (infant-grade silicone refinement process).
As our CEO often reminds the team when new team members are introduced to our IP portfolio: a patent is not a product. A patent is an investment in a specific technical solution to a specific manufacturing problem. Most of our patents address narrow but commercially important questions: How do you maintain antibacterial properties in a silicone compound without toxic additives? How do you improve uniformity in a multi-step silicone processing chain? How do you bond silicone to a specific substrate without delamination under thermal cycling?
The reason 300+ patents matter for buyers is not braggadocio — it is a signal about organizational culture. Companies that patent solutions are companies that invest in solving problems rather than accepting them. When our process produces a quality issue, our technical team's reflex is to understand it at the molecular level and develop a systemic fix, not to tighten an inspection criterion and move on. Patent CN113650314A exists because we were unsatisfied with the industry standard approach to antimicrobial silicone (which typically involves silver-ion additives) and developed an alternative formulation that works without added metals. That is the kind of problem-solving culture our patent portfolio reflects.
For buyers, the practical implication is this: when you bring us a product development challenge, we are not looking it up in a supplier catalog. We are drawing on 21 years of documented, patented, production-validated solutions.
We have been in this industry long enough to have seen everything that goes wrong. Here are the things that are widely known inside factories and rarely disclosed to buyers.
Compound substitution is more common than you think. A factory quotes a price based on Wacker 3003/60A compound (or equivalent premium grade). The purchase order is signed. Production runs on a domestically-sourced compound that costs 40% less and has no third-party migration test. The finished product looks identical. The LFGB or FDA test is not run because "it is the same silicone." It is not the same. Demand lot-specific compound documentation on every production run.
Mold maintenance is the first thing cut when factories are under margin pressure. A mold that should be polished every 30,000 shots gets run to 60,000 without maintenance because the press is committed to another order and the tooling team is short-staffed. Flash increases gradually — not dramatically. Each individual batch passes. The aggregate result is parting lines that grow wider over 50,000 units without anyone formally rejecting a lot. Demand mold maintenance logs on request.
Post-cure shortcuts are invisible. A peroxide-cured HTV part that has not been fully post-cured carries elevated residual volatile content — acetic acid and other by-products. The part looks, smells, and tests dimensionally correct. It may even pass a quick migration test at the shorter time-temperature conditions a factory uses to self-check. It will fail a rigorous LFGB test at elevated temperature with fatty simulants. Post-cure documentation — oven temperature records, batch records showing time in oven — should be available for every peroxide-cure production run.
"Factory" and "trading company" are not always clearly different. We have covered this elsewhere. From the inside, the clearest signal is this: a factory that can discuss your product's DFM challenges in specific engineering terms — parting line, shut-off angle, vulcanization curve, AQL — is almost certainly a real factory. A company that redirects every technical question to a "technical team" that is unavailable is almost certainly a trading company.
Rush orders compromise quality. Not because factories are malicious — because physics and process control take time. Silicone cure cycles cannot be accelerated beyond the material's kinetics without creating residual stress. HTV post-cure cannot be shortened without leaving residual volatiles. Flash trimming cannot be rushed without increased operator error rates. When we say 20–35 days mass production lead time, it is not padding. It is the minimum time to produce consistently to specification.
Metrics that reflect where we are in 2026, drawn from our actual production records and operational data.
4,500 m²: total production floor area. Not a number we invented — it is the registered leaseholder square footage across our Shenzhen buildings.
100+ employees: full-time headcount. Our average production floor tenure is 6.4 years. That retention rate is deliberate — experienced operators are a competitive asset.
20+ LSR injection machines: currently active. We have added 4 machines in the last 3 years as baby and food-grade product demand has grown.
30,000 units/day: committed daily output capacity across both LSR and HTV lines. This is a conservative number — actual peak capacity is higher, but we commit to what we consistently deliver.
300+ patents: granted across our portfolio, including 3 core invention patents.
50+ countries: active export destinations as of 2025, spanning North America, EU, UK, Australia, Japan, and emerging markets in Southeast Asia and the Middle East.
MOQ 300 units: for OEM projects sharing our standard mold base. We set this number specifically to be accessible to brand-stage buyers, not just enterprise accounts.
7-day sample lead time: standard for projects where a mold exists or is adapted from our standard base. DFM-to-sample in 7 days is not a target — it is our standard operating procedure.
21 years: from our founding in 2005. Every year, we have shipped product. Every year, we have learned something. Most years, we have learned something the hard way.
Every new factory has enthusiasm. Many have capable equipment. Some have talented engineers. What a 3-year factory cannot have is accumulated data — the production records, compound performance histories, mold maintenance logs, customer complaint patterns, and regulatory change responses that span hundreds of product launches across dozens of market segments.
When a buyer comes to us with a product that has failed EN 71-3 chemical migration at a previous supplier, we almost always know why before we see the test report. Either the compound was not properly specified, the colorant was not verified against EN 71-3 limits, or the test was run at the wrong condition. Our prior experience with the failure mode makes our diagnosis fast and our solution reliable.
When a buyer's retail customer changes its compliance requirements — as Walmart did with its Product Safety and Compliance portal in 2023, adding new documentation requirements — we notify all affected accounts proactively and have documentation updates ready before buyers need to respond to their retail buyer. We know because we monitor the changes. We monitor because we have been doing this long enough to know that regulatory environments evolve continuously.
When a buyer says "I need this in Shore A 45, FDA 21 CFR 177.2600 compliant, Pantone 485 red, within Delta E 1.0, 30,000 units, 25 days" — we know from our compound performance database exactly which compound family matches that specification, which colorant achieves Pantone 485 within tolerance, and which machine configuration produces the most consistent color across a 30,000-unit run. We do not figure this out. We look it up.
That is what 21 years of manufacturing data, systematically organized, looks like in practice.
Perfection is not a realistic claim from any manufacturer, and claims of perfection should be treated as a red flag rather than a selling point. Here is an honest account of where we fail and how we respond.
First articles that miss tolerance. Our T1 sample acceptance rate is approximately 78% on the first round for new OEM designs. That means roughly 22% of new molds require a T2 revision before the design is locked. The causes are almost always one of three things: colorant-to-compound interaction that shifts properties slightly from calculated, parting line witness mark visibility on a surface where the buyer's aesthetic tolerance turns out to be tighter than the drawing specified, or dimensional deviation on a feature with a tight fit to a mating component. We handle T2 rounds within 5 working days and at no additional charge for design-driven deviations.
Delivery delays. On average across 2024, approximately 4.8% of our mass production shipments were delayed by 3 or more working days versus committed date. Causes were split roughly equally between compound supply delays (our suppliers' lead times are not always consistent) and production capacity constraints during peak periods (Q3 and Q4 holiday run-up). We notify buyers of potential delays a minimum of 10 days before the original ship date where possible, and we have not had a customs-documented late shipment without prior notification in over three years.
Color variation between runs. Our colorimeter Delta E data shows run-to-run variation within Delta E 1.5 for 94% of production orders; 6% of orders have at least one production lot with Delta E between 1.5 and 3.0. For most buyers, Delta E under 2.0 is visually imperceptible. For cosmetics-grade or premium retail products requiring tighter color consistency, we recommend retaining reference plaques at our facility and requiring visual approval of the first production batch before bulk shipment.
We do not hide these numbers. A factory that claims zero defect rates on first articles is a factory that does not measure them properly.
As our CEO often reminds the team as visitors are leaving the factory floor: "Show them everything. If what they see makes them uncomfortable, that is information. If what they see gives them confidence, that is the relationship."
We have been in Shenzhen since 2005. We have shipped silicone products to buyers in 50+ countries. We have survived the 2008 financial crisis, two significant compound price spikes, COVID-19 supply chain disruption, and multiple rounds of increasingly stringent regulatory requirements. We are still here because we make the product we say we make, we test it the way we say we test it, and we tell buyers the truth when things do not go as planned.
This article is the longest thing on our website. It should be — it is 21 years of what we know.
If you read this and think we might be the right partner for your silicone program, reach out. If you read this and think we are not the right partner, we appreciate your time. The right factory-buyer relationships are built on fit, not on sales pressure.
Learn more about our history and founding
Visit our about us page
Explore our factory overview
Review our OEM/ODM capabilities
See our certifications
Email: yfxy@mymitour.com
Phone: +86 199 2529 4106
Mitour Silicone was founded in 2005 and is based in Shenzhen, China. We have operated continuously from the same city for 21 years, growing from a small HTV compression operation to a 4,500 m² facility with 100+ employees, 20+ LSR injection machines, and 30,000 units/day capacity.
Solid-to-Liquid Refinement Technology (a proprietary process developed in-house by Mitour Silicone) is a processing pathway that converts carefully selected solid silicone raw material into a refined compound with LSR-equivalent purity. It is protected by patent CN114015239A. For infant-grade applications, it delivers reduced volatile content (D4/D5 cyclosiloxanes below 0.5 ppm), improved catalyst distribution uniformity, and simplified material traceability.
Patent CN113650314A covers Mitour Silicone's antimicrobial silicone composition — a formulation that provides antibacterial properties without relying on silver-ion additives or other metal-based antimicrobials. It applies to baby products, pet accessories, and any silicone application where inherent surface hygiene is important. Buyers can request the technical overview and associated efficacy test documentation.
In spring 2019, a Walmart-approved audit body conducted a 147-point SMETA 4-Pillar and manufacturing capability audit at our facility. We received one non-conformity (a color mixing log missing compound lot cross-references), corrected within 14 days. Since then, no subsequent Walmart audit has issued a non-conformity. Walmart and Target became formal supply clients in 2019.
For first article revisions: T2 rounds are completed within 5 working days at no charge for design-driven deviations. For delivery delays: we notify buyers a minimum of 10 days before the original ship date where possible. Our 2024 data shows 4.8% of shipments were delayed 3+ working days, with all delays communicated in advance. We do not claim perfection; we claim transparency.
We currently export to 50+ countries spanning North America, EU, UK, Australia, Japan, Southeast Asia, and the Middle East. Our compliance documentation covers FDA 21 CFR 177.2600 (US), LFGB §30/31 (EU), EN 71 (EU toys), BSCI/SEDEX (social compliance), and REACH SVHC (EU chemicals). We can support market-specific compliance requirements for additional markets on request.
MOQ is 300 units for OEM projects sharing our standard mold base (project-dependent for fully custom geometry). Standard sample lead time is 7 days from mold approval for existing tooling. Mass production lead time is 20–35 days depending on product complexity and order volume.
Email yfxy@mymitour.com with three pieces of information: (1) what the product is and who uses it, (2) your target annual volume, and (3) any regulatory market you need to comply with (US FDA, EU LFGB, EN 71, etc.). Our team responds within one business day with a DFM feasibility assessment and preliminary cost structure. We do not require NDAs to start a technical conversation, but we respect them if you prefer one.
Twenty-one years in this industry has taught us one reliable predictor of good buyer-factory relationships: buyers who ask specific, informed questions get specific, informed answers — and those conversations become the foundation of partnerships that last.
This article is our attempt to equip you with the questions worth asking, whether you ask them of us or of any other silicone manufacturer you are evaluating.
Email: yfxy@mymitour.com
Phone: +86 199 2529 4106
Peter Cui | 21 years of silicone manufacturing experience | 4,500 m² Shenzhen facility | Walmart-, Target-, and Disney-approved supplier | Contact: yfxy@mymitour.com
