Has anyone quantified area bias from syringe filter material for moderately hydrophobic analytes? In 50:50 acetonitrile:water I’m seeing 4–6% lower peak area for cinnamaldehyde after PVDF 0.22 µm vs PTFE filtration on an Agilent 1260 (UV 290 nm, n=5, carryover <0.05%), which reads like membrane adsorption rather than system variance — does a solvent pre-rinse normalize recovery for you, or do you just switch membranes?
PVDF can be a bit grabby — — so I condition: 2–[redacted] neat ACN through the 0.22 µm, then 2–[redacted] of your 50:50, and discard the first about [redacted]; that gets cinnamaldehyde within about 1–2% of PTFE for me on a 1260. Alternatively, do a quick ‘sacrificial’ pass with a dilute standard or use low‑bind PVDF; PTFE is fine too, just be sure it’s truly wetted (less of an issue at 50:50). Good overview here: https://www.sigmaaldrich.com/US/en/technical-documents/technical-article/analytical-chemistry/sample-preparation/choosing-syringe-filters — did your “pre-rinse” include neat ACN?
I saw the same “4–6% lower peak area” on PVDF in 50:50, and what fixed it was a sacrificial pass: push about [redacted] of a slightly higher‑conc cinnamaldehyde standard through the 0.22 µm to waste, then run your sample on that same filter on the 1260. PTFE didn’t need it for me, but if volume is tight, PES behaved closer to PTFE for aromatics; can you spare that [redacted] per prep?
Building on @carla_jones34: a tiny IPA pre‑wet (≈50–100 µL) of the PVDF, then a quick flush with your 50:50 and no dwell time, knocked my cinnamaldehyde drop to about 1% — PVDF can be a little Velcro for aromatics. If that doesn’t normalize, RC membranes have been the least sticky for me; what filter brand/diameter are you on?
But agree with @juliaf67 on conditioning — what really helped me was ditching the [redacted] PP syringe for a glass luer‑lock, which dropped the PVDF loss on cinnamaldehyde in 50:50 from about 5% to about 2%; are you on plastic syringes?
, PVDF + cinnamaldehyde in “50:50 acetonitrile:water” loves to soak up a few percent. A [redacted] neat ACN pre‑rinse followed by [redacted] of your 50:50 — discard both — cut my PVDF 0.22 µm bias on a 1260 to about 0.5%, and hydrophilized PVDF behaved similarly. @carla_jones34 is right on quick handling; also check for glycerin‑prewet housings and, at low µg/mL, do a 100–200 µL sacrificial sample push to waste to saturate sites.
If you want to stick with PVDF, pre-load it with a small ‘sacrificial’ shot of your mobile phase spiked about 5× in analyte (0.5–[redacted]), discard, then run the sample — feed the filter first; this cut my loss to about 1% in a similar setup. Minor caveat: push at a steady, faster rate; slow dripping made adsorption worse for me. @emilyross92, did you see a flow-rate effect too?
I saw the same ‘4–6%’ on an Agilent 1260 at 290 nm with 0.22 µm PVDF — pushing through quickly and discarding the first about 300 µL of the actual sample consistently pulled the bias under 1% for cinnamaldehyde. What diameter are you using — 13 or 25 mm — since smaller area helped a bit for me as well.
With PVDF, a quick 100 µL ‘ACN pre-wet’ followed by about 500 µL of your 50:50 before the sample cut my cinnamaldehyde drop from about 5% to about 1–2% on an Agilent 1260. Small caveat: if you’re working near LOQ, switching to 13 mm PVDF trimmed surface contact and tightened recovery a bit.
Surface binding screams hydrophobic PVDF in 50:50 ACN:water — check the spec says “hydrophilic PVDF”; if not, swap to RC 0.22 µm, which gave me about 99% recovery for cinnamaldehyde at 290 nm. If you must stay PVDF, use the hydrophilic variant and keep the push continuous to limit contact time. What brand/model is your 0.22 µm PVDF?
Quick check: when I switched from polypropylene (PP) to glass syringes, the cinnamaldehyde drop with PVDF shrank from about 5% to about 1–2%, so the syringe barrel was part of the loss. Are you on plastic syringes? If you stay on PVDF, a brief ‘30 s pre-soak with your mobile phase’ helped me, but glass made the bigger difference.
And one thing that fixed this for me was a ‘saturation rinse’: push 1–2 void volumes of a quick spike (same matrix, about 5× cinnamaldehyde), discard, then filter the actual sample — PVDF bias dropped to about 1% vs PTFE. To prove it’s adsorption, wash the used disc with neat ACN and re‑inject; mine coughed up the “missing” area like a vending machine refund, and faster, consistent plunger speed also helped by cutting contact time, so the extraction check should tell you quickly.
I’d pre-wet the PVDF with about [redacted] IPA, then flush 1–[redacted] of your 50:50 and discard before the sample — on my 1260 at 290 nm this shrank a similar 4–6% gap to about 1%. If it still drifts, try PES or share the brand/hydrophilic spec you’ve got; @carla_jones34’s approach works but this avoids spiking.
In my hands the about 5% drop with PVDF at 50:50 turned out to be the “glass fiber prefilter” scavenging cinnamaldehyde, not the PVDF itself. Switching to the same brand’s PVDF without the GF layer pulled recovery to within about 1% of PTFE on a 1260 with no other changes. Are your PVDF units the GF-prefiltered type?
Try a quick fraction test: collect the first [redacted] after PVDF, then the next [redacted]; if the second recovers, it’s adsorption breakthrough, so push one void volume fast and don’t “rest” the plunger — don’t let cinnamaldehyde make friends with the membrane. If both fractions stay low, suspect the housing or wetting additives rather than the membrane and either switch to the same model in PTFE or jump to RC; @aronb56, do you recall the housing polymer?