Every chemist knows C V = C V .
So why do dilutions still fail?
Because the formula isn't the problem. The execution is.
Three traps I see in every lab including mine:
Trap 1: The unit trap.
A stock at 100 g/mL. You need 100 ng/mL. That's a 1000 dilution, not 100 . One missed " " and your dose is off by an order of magnitude. The formula is correct. The concentration isn't.
Trap 2: The "unpipettable" dilution.
C V =C V says: take 1 L of a 10 mM stock, add 999 L buffer. Mathematically perfect. Physically insane you're trusting a P10's lowest range for your entire experiment.
Trap 3: The wrong pipette.
A P200 delivers ~1-3% accuracy at the top of its range but 3-5% near the bottom. If your "100 L" is actually 95, your dilution is 5% off before you start.
Here's what I love about SciTuu's dilution calculator (scituu.com, free, runs in your browser):
C V = C V Stock: 100 M
Target: 10 M
Final: 1,000 L Pipette 100 L stock + 900 L diluent Feasible with a P200 pipette
It doesn't just give you the number. It tells you whether the plan is actually executable and with which pipette.
That tiny "Feasible with a P200" line saves more experiments than any spreadsheet ever will. Because it forces the question: is this dilution physically sane, or just mathematically correct?
Chemistry is a bench sport. The math is the easy part. The pipette is where experiments actually die.
What's the worst dilution mistake you've seen or made? Asking for a lab that's definitely not judging.
#Chemistry #Dilution #LabLife #WetLab #Pipetting #SciTuu #LabAutomation #ChemistryMemes
Hi Product Hunt — Kenny here, maker of SciTuu. 👋
I built SciTuu because routine wet-lab work still jumps between spreadsheets, instrument files, notebooks and manual handoffs.
SciTuu is a free, privacy-first toolkit for wet-lab planning. It now combines 29 deterministic calculators with Data Inbox, Instrument Inbox, guided workflows and Experiment Packet 2.0 exports.
You can start with qPCR, ELISA, cell culture, Gibson/clone QC, plate planning or Opentrons; import CSV/XLSX/JSON, NanoDrop or plate-reader data; map it locally; and export reviewable records for Benchling/LabArchives, Slack/Teams, cloud storage, AI4S/standards metadata or Remote MCP.
Core calculations run in the browser. Accounts and integrations are optional. Research use only.
I’d love feedback on:
• Which instrument file or lab workflow should we support next?
• Would Data Inbox + Experiment Packet fit how your team reviews experiments?
• Which handoff matters most: ELN/LIMS, Slack/Teams, Opentrons or AI/agent workflows?