Contaminated laboratory water

[bisnettrj2]

Hello all:

I had a discussion with Uwe Neue regarding Waters SPEs, and it got me thinking (which is usually not a good thing, I'm a theoretical can-of-worms kind of guy) - how does your lab qualify its water source? When you say you are using "reagent grade" water, how do you "prove" that? When you use a distilled water source or a Millipore-type purifier, how do you qualify that water as interferent and contaminant free? My question stems not only from my posting and subsequent discussion with Uwe, but also from a recent discussion with a colleague regarding his use of an outside water source for analysis of a method blank for EPA 1668 analysis, because our lab water and distilled water were giving him hits in the method blanks. If anyone can give me some solid reasons and practices in terms of qualifying water sources, I would love to change the culture in my lab and bring some better practices into use.

[unmgvar]

bisnettrj2,

following your discussion with Uwe regarding the SPE, and based only on this right now, i would first define the type of contamination we are faced with and the impact it has for you.

the subject can be really broad so i will give a very simple and narrow example of what i mean, and yes it can and surely is more complicated then this.

alright let say i want to use my water for chromatography, so i have to asses the risks involved in using it.
let's say that water can cause 3 things:
high drift effects
high noise levels
compound contamination.
how to resolve this?
for the first 2 follow a test criteria kind of like what HPLC vendors do for their detectors.

for the last one doing UV scan in a spectrophotometer, against a high purity grade water source from 190-800 nm would for example show you if your water is contaminated with chormatographic impurities that will affect your chromatograhy.

[HW Mueller]

Other type of detectors might be quite handy in detecting "dirt". However, the stuff might be so low that no detector sees it, but it may accumulate on the column. As discussed many times, a gradient or step change in solvent might reveal this (in other words, you have to look for this sort of thing with the chromatography, another method is mentioned below). With detectors like UV one may also have the problem that there can be substances which do not have a visible maximum in the wavelength range of interest (scattering, refractive index phenomena, even substances with max below 190nm).
In my own laboratory work I am confined to isocratic work or step changes of mobile phase. That´s why I concocted the device described the other day, essentially a prep. column converted to a gas pressure mobile phase delivery system. I check mostly the baseline of water, namely of every 2.5 L bottle I collected (Milli Q), always comparing it to a previously optained H2O. Sometimes I compare baselines using this device between column and no column (careful, one has to compare the column under equilibrium, this tells me whether there is a considerable column bleed and or whether something gets adsorbed).

[krickos]

Hi

Some feedback to the thread creator.

a: a link to an article a bit "sponsered " by a company it seems
http://www.millipore.com/bibliography.n ... 95EN00.pdf

b. What we use water wise. Locally produced LC water via osmosis etc (not Millipore), ie >18Mohm and in line TOC reader (high/low/trace).
Each production unit is tested regulary (much in the same way as Purefied water for pharmaceutial use (Ph Eur/USP) includes microbiology).

Generally the HPLC water production units works well with regard to quality BUT we have issues with the time it sometimes takes to get down to low TOC levels.

[Bruce Hamilton]

It's important to ensure that your water is fit for purpose and, if in a regulatory environment, production and testing of the water should match the specification. Millipore has plenty of free advice on their website, and offer a diverse range of Milli-Q systems for labs.

Generally, for chromatography, the specifications are fairly simple, eg Milli-Q, tested to ACS/ASTM, EP, USP etc. The issue of trace contaminants really depends on what the water is being used for. The limits are different if you are performing Ion Chromatography, trace inorganics, trace organics, etc.

TOC is increasingly being specified for pharmaceutical HPLC applications, however, as Krickos noted, some systems can take a lot of flushing/recirculation to reduce the value, and if you are performing some analyses, eg halogenated pesticides, TOC may not be appropriate. If you examine the specification for bottled pure water, you will often find they offer different Ultrapure grades for different applications.

For chromatography they may perform UV gradient testing, UV spectrum, fluorescence, residue, and bug count/specific conductance at time of bottling.

UV gradient testing using a known good solvent ( I prefer acetonitrile - the method is usually listed in the monograph ) is a simple technique that can provide baseline data to help identify the origin of peaks and high background in the future.

Lab units, such as Milli-Q configured for chromatography, are maintained and serviced according the the supplier, and the product water tested according tro the regulator. The qualify of the feed water into the unit is one of the main determinants of final water quality.

I assume your laboratory water supply has been "fit for purpose" until now, so applying expensive control criteria and buying new purification systems may not be justified, but it's certainly worth raising with your management, and performing a risk assessment, even if no change is planned.

For problem solving, it's convenient to have two sources of pure laboratory water to help exclude water issues.

It's also really important to ensure that the containers the water is being dispensed into are also of suitable quality and cleanliness. Pure water is a universal aggressive solvent - ask any corrosion engineer.