Weird science: Research surprises and accidental discoveries

Scientific research and lab work seldom go exactly as planned. Despite impeccable preparation and precise planning, there are too many variables to perfectly project results (they’re called experiments for a reason). Quartzy’s resident scientists share the weirdest lab and research surprises they’ve ever experienced—some of which led to accidental and valuable discoveries.

George, Marketing
Surprise: Back when I worked in a failure-analysis lab, I did most of the testing that involved chemistry (being the only analytical chemist). There was a particular chemical synthesis we were trying to recreate that kept going wrong, and no one could figure out why. Eventually, we brought the customer in to observe our procedure, and he, too, was baffled. The next day, while working on the same problem, one of our new lab techs took off his gloves due to a “tingly feeling.” His hands showed signs of light acid exposure, which was odd since we were not working with any acids. I asked him when the last time he changed his gloves was, and it was longer than it should have been, considering we were working with dichloromethane (dichloromethane penetrates nitrile gloves). Still, DCM should not cause chemical burns—though it will cause other nasty things. Since we were in an analytical chemistry lab, we did some analysis on the DCM and found a significant enough amount of sulfuric acid present. It turns out we got a bad production batch of DCM that the supplier’s QC amazingly did not catch. This was probably the first and only time I got a “bad” solvent, but it quickly explained why our experiment was failing. After settling things with our supplier and getting new DCM, the synthesis worked.
Discovery: Poor quality DCM is sometimes industrially processed with sulfuric acid. When things go wrong, don’t forget to check your chemical purity and certificates of analysis.

Aline, Life Science Product Specialist
Surprise: When I was interning at a winery, one of my jobs was to go sample vineyards at the butt crack of dawn. I used to drive two hours to Mendocino County to sample various blocks and rows within the vineyard. Our winery used my samples to decide when to pick the fruit. I helped make the call on one of the blocks, we got the fruit, and it was several g/L off of the sampled sugar value. I swore up and down that I had taken good, random samples from several nodes of the block and couldn’t rectify why the harvested fruit was so different from the sample.
Discovery: The next day, I went back out to sample some more blocks and revisited my “problem block.” Turns out, the harvesting company had screwed up and taken several rows outside of the block that they weren’t supposed to. I took some photos and told on the harvesting company to the winemakers. If your sampling protocol is optimized for randomization and equal representation, then trust the data!

Dylan, Customer Success Associate
Surprise: When collecting and analyzing data for my thesis about making a linkage map (think a rough scaffold of an organism’s chromosomes), I was doing a lot of next-generation sequencing of a particular Joshua tree and its offspring: seedlings I grew in the lab. When I got to the step to infer the “father” of the tree via a fancy program, we discovered that there was no father—the tree had actually self-pollinated!
Discovery: Up until this point, no one knew if Joshua trees could self-pollinate or not! Not only was this an interesting finding, it also made the downstream efforts a little simpler for the programs I needed to run to actually make the linkage map.

Melissa, Life Science Product Specialist
Surprise 1: A few years ago, I was working for a reference lab that had a “sister lab” in India. I was sent there to validate assays and train their lab personnel on a new panel of PCR/qPCR assays. Everything was going great until we tried to validate the most anticipated assay of the panel. It just wouldn’t work. We tried three iterations of the experiment and nothing worked. I was starting to look bad, and had to explain to my bosses why this wasn’t working. My only explanation was that the assay kit had to be faulty.
Discovery 1: The kit manufacturer had switched its running buffer (and dye), which was not compatible with some qPCR machines. They had to redesign their entire line of that assay. Whoops!

Surprise 2: During my master’s thesis research, I was testing a bunch of genes to try to identify which apoptotic pathway was utilized for our recombinant peptide to induce cell death in melanoma cells. We had assumed that a traditional, fairly well-known pathway was used. My experiment was meant to sort of confirm information we already suspected.
Discovery 2: Genes were amplified that we did not expect! We decided that there must be some crossover of apoptotic pathways. My old lab will be able to use this as a jumping off point for a whole new set of experiments.

Do you have any personal stories of scientific surprise? Send us an email!

Quartzy is the world’s No. 1 lab management platform. We help scientists easily organize orders, manage inventory, and save money. We’re free and always will be. Visit or reach out at

Interested in writing for The Q? Send us an email!

Greg Schindler

Greg has a BA from Stanford (English/Football) and MS from Oregon (Journalism). He's our Director of Marketing and Pastries.

Leave a Reply

Your email address will not be published. Required fields are marked *