Bookmark and Share

From: Food Quality & Safety magazine, June/July 2005

Choosing a Laboratory Water System

A water production system supporting analytical methods in food laboratories needs to reflect a firm's use of different water quality grades and workflow.

by Grec Crescenzi and Joseph Plurad

Food analysis laboratory needs for purified water tend to be modest in quantity but demanding in quality. Food analytical methods call for two general water quality grades, pure and ultrapure. To meet purified water needs, cost-conscious laboratories consuming up to 15 liters of water per day must choose from among several options:

  • Purchasing and maintaining separate purification systems for each grade of water used;
  • Acquiring a single, ultrapure water system and absorb the capital and maintenance costs of producing very high-quality water for less-than-critical applications;
  • Buying bottled water that roughly corresponds to pure and ultrapure standards;
  • Obtaining an all-purpose water system that supplies water on demand at the appropriate purity level for every assay.

The decision ultimately is based on quality, convenience and cost. Generally, small in size, food laboratories often combine R&D with quality-related work. The assays lab techs conduct can include quality testing of raw materials and finished products. The goal is to determine food safety requirements, nutritional analysis, quality assurance/quality control and product development.

Water quality needs to be considered for optimal results. Operations requiring purified water include most analytical methods (titration, HPLC, ion analysis, colorimetric assays), microbiological tests, quality control, growth media, washing and rinses, column analytical and preparative chromatography, etc.

The two major classes of laboratory water as pure and ultrapure. Pure water has been distilled, deionized or treated by reverse osmosis. It is used in non-critical applications, where detection limits are high or in situations in which contaminants from water purity fall well within the error limits of the method used.

Ultrapure water takes pure water and adds purification steps that kill adventitious pathogens as well as removes trace ions and organics. Not every application that calls for ultrapure water demands removal of all remaining contaminants up to ultrapure standards. Specific analytical methods may only require depletion of certain ions, bacterial products or organics.

The need to remove organic contaminants is the most common reason for specifying ultrapure water. Organics interfere with analytical HPLC by altering peak resolution and integration, introducing ghost peaks, and affecting stationary phase chemical selectivity. About 70 and 80 percent of HPLC performance problems are directly attributable to water quality. For analytical HPLC, organic contaminants easily can achieve on-column concentrations equal to those of target analytes. In preparative chromatography, organics may concentrate on columns and co-elute with product.

Analytical-grade ultrapure water should comply with ASTM Specification D1193, which stipulates that water be freshly drawn and used within 8 hours of production. ASTM D1193 also specifies acceptable levels for dissolved total organic carbon (TOC) at less than 50 ppb, and microbiological contamination to less than 10 colony-forming units per liter.

Standards for ultrapure water are significantly higher than for pure water, specifically with respect to removal of organic contaminants, which include free-standing hydrocarbons, halogenated organics, detergents and bacterial products.

Ultraviolet-mediated photo-oxidation is the most efficient, cost-effective means of achieving removal of environmental and pathogen-related organics. UV photo-oxidation employs a dual-wavelength, low-pressure mercury UV lamp in quartz sleeves and dual-wavelength (185 and 254 nm) irradiation. This wavelength combination generates hydroxyl radicals from dissolved oxygen and water. Hydroxyl and secondary free radicals react with and break up small organic molecules, the end products of which are water and CO2.

Difficult Choices

Bottled water is perhaps the least cost-effective solution. Ultrapure HPLC-grade bottled water costs close to $20 per liter - far higher than the cost of water produced by an ultrapure water system.

A jug of distilled water costs between $1 to $2.50 per gallon in supermarkets and pharmacies. While this water is certainly of very high quality for ordinary household chores, using it in a laboratory setting is not advisable. It's impossible to determine the pedigree of bottled water that is not reagent grade. Plastic containers used to store jug water may leach chemicals into water. And once the bottle is opened all bets are off since water absorbs gases and chemicals present in laboratory air.

Reagent-grade bottled water is rated to its quality specification at the time of bottling or, in some cases, when the bottle is opened. Once the manufacturer's seal is broken water begins absorbing CO2 present in the headspace of the container. Water stored in improperly-sealed containers can rapidly pick up amines, aldehydes, alcohols and other common laboratory chemicals in the air.

Laboratory managers rightly bristle at the prospect of purchasing and maintaining two water systems - one for ultrapure water and a second for routine operations and/or to feed the ultrapure system. Single-unit, standalone systems that deliver both grades of water provide higher value and convenience.

A cost-effective water system, regardless of the quality of its output, should use the least expensive feedstock water possible, which is normally tap water. Systems that require a separate source of purified water to feed ultra-purification add an additional layer of cost consumption.

Nevertheless, some food laboratories, whether through custom or other circumstances, find themselves with two separate systems.

Distillation, deionization and reverse osmosis are typical methods for generating pure-grade laboratory water. Stills employ holding tanks in which water sits for hours or days, whereas deionizers feed directly into the ultrapure stage. Carbon dioxide uptake during storage is known to lower product water pH, which affects acid-base titrations and other types of analysis. Blanks effectively will cancel the effect of pH and other chemical changes provided blank readings are obtained within minutes of analytical runs, an unlikely scenario for busy analytical laboratories.

Distilled water is considered pure because most of the particulates, pathogens, or ions present in the feedstock water end up in the distillate. However, ions, particles, and colloids may sweep into the vapor phase and condense along with product water. Although distilled water is pure enough for routine laboratory work, it is best used immediately after producing it, since the relatively high condensate temperature promotes growth of microorganisms.

Atmospheric carbon dioxide easily dissolves in any water on storage. Given the shortcomings of distilled water, it cannot be recommended as feedstock for ultrapure water systems.

Deionization, another common route to pure water and feedstock for ultrapure water generation, also suffers from drawbacks. Although acceptable as a source of "as-is" laboratory-grade water, deionized water does not efficiently remove organic contaminants from common feedstock water. Considering that deionization commonly uses tap water as its source, and most tap waters contain relatively high levels of organic contaminants, deionization is not the most appropriate choice for feeding an ultrapure water system for food laboratories.

By contrast reverse osmosis removes dissolved and suspended species such as particles, bacteria, viruses and macromolecules (e.g. sugars, proteins) with molecular weights above 150 to 250 daltons by simple size exclusion. Microorganism proliferation is far less likely after reverse osmosis, which occurs at ambient temperature. Reverse osmosis removes 95 percent or more of all contaminants likely to be present in tap water. In so doing, reverse osmosis generates pure-grade water suitable for routine laboratory use and as feedstock for ultrapure water production.

Most dedicated ultrapure water systems are not designed to use tap water as feedstock, necessitating the purchase of additional purified water systems or the use of expensive bottled reagent water. Some ultrapure water systems employ built-in electrodeionization, deionization or reverse osmosis pretreatment systems. Others require a stand-alone pure water source to feed into it. Due to poor synchronization of output between purified and ultrapure systems, most dual-system arrangements require a holding tank to store purified water.

Dedicated systems that utilize potable water sources tend to be incapable of generating intermediate-grade purified water. As a result, end-users must bear the cost of using ultrapure water for routine operations.

Laboratories with mixed quality water requirements would do well, therefore, to source water systems that reflect water usage. Incurring the unnecessary cost of ultrapure water for less-critical operations should be as unthinkable as compromising on quality for highly sensitive assays.


A water production system supporting analytical methods in food laboratories needs to reflect a firm's use of different water quality grades and workflow. For most food laboratories that means providing ultrapure water for highly-sensitive assays, as well as lower-cost pure-grade water for less critical operations. Labs that use both pure and ultrapure waters may purchase two separate systems - one for each grade - and hope for the best after plumbing them together. However, most food analytical labs would do better to acquire a single system that produces both grades of water from inexpensive tap water.

The importance of "water-on-demand" cannot be over-stressed for analytical laboratories, especially those performing sensitive assays that may be compromised by low-quality water. On-demand sources of pure and ultrapure water assure that water quality will not change over time due to introduction of gases, particles, or solutes. Single-source systems also provide the greatest level of convenience, and in the long run economy as well.

Greg Crescenzi is business development manager and Joseph Plurad is marketing communications manager for Millipore Corp.'s Bioscience Division (Danvers, Mass.). Reach Crescnezi at 978-762-5312 or Plurad can be reached at 978-762-5370 or



Current Issue

Current Issue

April/May 2014

Site Search

Site Navigation