In March 2024, I sat down to review a set of product information pages for the Mitsubishi Chemical Group Corporation website migration. I thought it would take about two days. It took almost a month, and it ended up changing the questions I ask about every material grade we publish.
I'm not the product developer. I'm the quality manager who checks the final wording before customers see it. That means I spend a lot of time asking what a sentence actually means. In one review cycle, I handled a Mitsubishi Chemical Kai'li golf shaft specification, an epoxy resin safety data sheet, a chromatography resin technical bulletin, and a cleaning procedure involving hydrochloric acid. That mix sounds odd until you remember that a materials company has to be precise in every direction at once.
Epoxy Resin Toxicity Is Not a One-Line Answer
The first near miss happened with an epoxy resin product page. A draft said 'non-toxic, safe to use in all applications.' I knew that sentence was wrong, but I had to explain why.
From the outside, 'epoxy resin toxicity' looks like one question: is it toxic or not? The reality is that cured and uncured epoxy resin are different materials for safety purposes. A fully cured epoxy is generally stable and inert. The uncured resin system, including hardeners and reactive components, is where the hazard usually lives. For many uncured epoxy systems, the SDS includes the GHS hazard statement H317: 'May cause an allergic skin reaction.'
Under the OSHA Hazard Communication Standard, hazard classification has to come from the specific SDS, not from a general statement about a resin family. The word 'non-toxic' invites someone to forget that the product has a cure state, a solvent content, and a handling requirement. The page we finally approved said: 'Consult the SDS before handling uncured material. Fully cured parts are generally inert, but machining them still requires dust control.' That is less flashy, and it is honest.
A Mitsubishi Chemical Kai'li Golf Shaft Taught Me To Check the Starting Material
The same folder included a product sheet for the Mitsubishi Chemical Kai'li shaft. To most golfers, it is a shaft that feels stiff and stable. To a quality reviewer, the interesting part is what happens before it becomes a shaft. Carbon fiber composite products start as prepreg: carbon fiber fabric with resin already in it. Before that prepreg is cured, the resin system has hazards.
I'm not saying a finished golf shaft creates the same handling risk as an open drum of resin. It doesn't. But if we market everything as 'advanced carbon material' without identifying the resin chemistry, we repeat the epoxy mistake with another name. The finished product and the raw material have to be described separately, because they usually do not share the same hazards. That was the turning point for me: a quality review should not only answer 'does it do what it says?' It also answers 'can someone use this message safely?'
When Someone Asks About the Leading Resins for Buffer Exchange Chromatography
Two weeks into the audit, a process development engineer sent a question: 'Which of the leading resins for buffer exchange chromatography should I compare for a scale-up from a 1 mL screening column to 100 L process column?'
I liked the question because it was specific. Buffer exchange is usually a size-exclusion step. The protein or large molecule moves through the packed bed faster than the small salts or reagents, so the sample ends up in the new buffer. The resins used for that step are often called desalting resins.
The important word in that sentence is not 'leading.' It is 'reproducible.' A supplier can call a resin leading because of market share, citations, or published data. A user still needs to know whether the packed column will behave the same way lot after lot. That is why a certificate of analysis matters. It is the evidence behind the performance claim.
If you are buying resin for a bioprocess, start by asking what the resin is supposed to do, what the contaminant is, what buffer conditions you need, and whether the supplier can provide data across multiple batches. Then ask whether the word 'best' is supported. Those are not the same question.
So, Are Hydrochloric Acid Fumes Dangerous?
The most important item in the review stack was not a polymer. It was a blog post draft about hydrochloric acid cleaning. The author had written an FAQ: 'Are hydrochloric acid fumes dangerous? No, because it is usually diluted before use.' I crossed that out before approving. That sentence could cause someone to work in an untreated tank and assume everything was fine.
Hydrochloric acid is water containing hydrogen chloride gas. Under certain conditions, that gas comes out of solution as a fume, mist, or vapor. It is a respiratory irritant at low levels.
In the US, OSHA sets a ceiling limit of 5 ppm for hydrogen chloride. NIOSH lists 50 ppm as immediately dangerous to life and health.
So the honest answer is: yes, hydrochloric acid fumes are dangerous if ventilation is poor, exposure is high enough, or the acid is mixed in a way that creates an aerosol. That does not mean dilute HCl cannot be used safely. It means safety depends on concentration, airflow, air monitoring, and PPE, not on a label that says 'no.'
The next sentence in that draft said 'always use in a well-ventilated area.' I changed it to 'use local exhaust ventilation and monitor the area when working in a confined space.' A drain in the floor is not ventilation.
What That Month Changed
After that file, I went back through every draft I had already approved. I was looking for one pattern: safety language that was too simple for the material involved. I found it more often than I wanted.
Now I run every public-facing product page through three tests:
- If it says 'non-toxic' or 'not dangerous,' which form of the material does that refer to? Is the user likely to handle that form or a different one?
- If it says 'leading' or 'superior,' what is the evidence? Market position can be true but irrelevant to application performance.
- If the SDS and the sales sheet disagree, which document is the customer told to follow?
Looking back, I should have started the project by asking each product steward about their biggest safety misconception. I didn't. I assumed the content team had already separated raw material hazards from finished product characteristics. That was my mistake.
Over the rest of 2024, I rejected about 12% of first-pass product documents because of unclear safety statements. That sounds like more work, and it was. It was also the useful kind of work.
A materials company earns trust not by saying 'no risk,' but by saying 'this is the risk, this is the form it applies to, and this is how you control it.' An informed customer asks better questions and makes faster decisions. That is why I see quality reviews as customer education, not as a bottleneck.