Detergent chemists analyze the difference in stain removers.

Properties such as “washes whiter than white”, “delivers maximum cleaning power”, and “gets rid of even the stubbornest stains (including blood)” have always been something of a Holy Grail for those of us who like to dress sharp. The detergent companies perpetually oblige our craving for ever more potent cleaning agents with “best ever” and “new, improved” brands year after year.

Until recently, however, putting a new spin on getting the most out of your washing machine has been something of a black art. But an original application of the latest analytical techniques is cleaning up the image of detergent science and leading to new products that really do make clothing come clean.

Gerard van Dalen, a researcher with Unilever’s Central Analytical Sciences group in Vlaardingen (Netherlands), compared several spectrometric techniques—X-ray fluorescence (XRF), X-ray photoelectron spectrometry (XPS), and Fourier transform infrared (FTIR)— to get to the heart of the wash and extract quantitative details about exactly how different materials, such as blood and soil, stain cloth. He has found that even the toughest stains can be analyzed down to the microgram-per-kilogram level with mass spectrometry (MS) and inductively coupled plasma–atomic emission spectroscopy (ICP-AES), and to the milligram level with XRF, XPS, and FTIR. This allows detergent scientists to quickly spot the difference between the old product and the new-and-improved version.

The Window Test
Optical reflectance measurements were, until recently, the detergent researcher’s standard and are familiar, in their consumer-test form, from countless TV ads showing the “before” and “after” shot on a bright sunny day. But optical reflectance does not distinguish between stain components, an aspect of soiling particularly important to those hoping to make a market-leading soap.

More information can, however, be obtained from a detailed application of modern analytical techniques. For instance, XRF, FTIR, MS, field emission scanning electron microscopy (FESEM), and XPS can reveal hidden details. XRF, explains van Dalen, can be used to determine how well a detergent removes soil particles by analyzing for elements such as silicon, aluminum, and iron. To determine protein levels in textiles, van Dalen applied some of these techniques. He chose to look at hemoglobin because bloodstains are particularly problematic, with a “burning-in effect” that makes them fairly permanent. Thus, any detergent that can get bloodstains out is certainly worth its salt.

Figure 1. Bloody Nuisance X-ray fluorescence (XRF) spectra of bloodstained and washed cloth. (Courtesy of Gerard van Dalen.)

Figure 2. Spectroscopic Bloodhounds. Fourier-transform infrared–photoacoustic spectroscopic (FTIR-PAS) profile of bloodstained and washed cloth. (Courtesy of Gerard van Dalen.)

Laundry, Ruddy Laundry
Proteins can be measured through their high nitrogen content using XRF (Figure 1 at right)XPS, and a nitrogen analyzer (N-analyzer), and through their amide bands as detected by FTIR (Figure 2 at right).These techniques offer information about the outer few micrometers (using XRF or FTIR) to nanometers (using XPS) of a cloth’s surface, explains van Dalen. The N-analyzer provides information about the bulk cloth. For studies of bloodstains, van Dalen points out that it is quite straightforward to test for iron and thereby quantify any hemoglobin present.

The van Dalen team used porcine blood from a local slaughterhouse and hemoglobin from Sigma-Aldrich Corp. as their test staining materials. Blank cloth swatches of cotton, polyester, and polyester-cotton blend were used as experimental garments. Creating the stains was as simple as preparing standard solutions and applying them with a pipette. Iron content was measured with XRF to provide a basis for relative hemoglobin content.

XRF, XPS, and FTIR measurements were carried out on the front side of the cloths and, according to van Dalen, indicated that the fabric was completely penetrated by the bloodstains to comparable levels. FESEM, however, revealed that, on the microscale, the blood was unevenly distributed over and within the fibers. Indeed, comparison of the nitrogen content of blood and hemoglobin stains demonstrated a logarithmic relationship between surface nitrogen—as revealed by XPS, XRF, and FTIR—and bulk nitrogen (N-analyzer).

According to van Dalen, the proteins in the cloths with a low quantity of blood or hemoglobin are concentrated on the outer layer of the cloth. He pointed out that although XRF was the best technique for following removal or build-up of protein on cloths because it is fast and precise, quantification was marred by this logarithmic relationship.

FTIR-attenuated total reflection (ATR) and FTIR–photoacoustic spectrometry (PAS) also produced a logarithmic relationship comparable to that obtained by XRF and XPS. “The log relationship is caused by the distribution of the blood in the sample,” explains van Dalen. “It cannot be avoided.” At low concentrations, the blood or hemoglobin is more concentrated on the outer layer and this, he adds, is observed when using surface techniques like XRF, XPS, and FTIR. It is not seen when using bulk techniques, such as nitrogen analysis, but nitrogen analysis is useful only at higher levels. “A big problem for surface techniques is quantification, that is, how to obtain homogeneous cloth standards, where bulk concentration equals concentration in the outer layer,” says van Dalen. However, XRF, XPS, and FTIR-PAS revealed a significant difference between cloths stained with 1% blood and hemoglobin compared with blank cloth because the protein is concentrated on the outer surface of the cloth. To the nitrogen analyzer, a difference at this concentration was invisible.

A Microscopic View
To confirm whether the stain remains within the cloth fibers or is a residual layer on the surface, van Dalen has also applied scanning electron microscopy (SEM) to the analysis of bloodstained cloths before and after washing. Obviously, if there is still deep penetration of the cloth after cleaning, then the test detergent is sadly lacking in power. “From the SEM images of residual bloodstains (cloth after washing), it seems the thin sheets of residual material are remnants of the bloodstain stretched out as self-supporting films, leading to the visibility of the residual bloodstain,” van Dalen says. So, even after washing, a light-yellow stain may remain on the cloth. “With the right detergent composition,” he adds, “this yellow stain can be removed.”

To examine the structure of the cloth samples more closely, van Dalen used FESEM. This allowed him to detect element-specific X-rays produced by the primary electron beam. Cotton fabric is composed of yarns spun from fibers, and SEM revealed the fibers of blank textile to be reasonably smooth. He found that the bloodstains on unwashed cloth are unevenly distributed over and between the fiber, and thick deposits and thin sheets stretch between fibers. After washing, the fibers look superficially clean, but thin sheets of residual bloodstain are apparent at higher magnification.

Technically Iron
More recently, van Dalen has used quantitative wavelength-dispersive X-ray fluorescence (WDXRF) spectrometry to measure the amount of iron on stained cloths quickly and accurately, without the need for sample pretreatment. He found a linear calibration line for an iron content range of 0–15,000 mg/kg. He suggested that the calibration line can be used to assess the staining and subsequent cleaning efficiency for cotton and polyester-cotton cloth stained with blood, clay, and iron oxide (to simulate rust). He emphasized that the measured concentrations have to be corrected when the thickness of these cloths differs from that of the calibration cloths, but that this is straightforward.

Then, van Dalen compared the WDXRF results with ICP-AES results. A detection limit for WDXRF was defined as the concentration that gives a net count rate equivalent to three times the standard deviation of the net count rate of a Teflon blank sample. For iron, this was calculated to be 4.8 mg/kg for the PW1404 and 2.6 mg/kg for the PW2404.

These are, van Dalen admits, higher than the theoretical limits, which he explains as being due to iron contaminants in the Teflon. But this can be improved upon by reusing only a limited number of carefully cleaned Teflon pieces and by stacking more layers of the cloth into the sample holder during the measurement and recalibrating. His results were within ±6% of the values obtained using the reference ICP-AES technique.

All the techniques can be useful, concludes van Dalen. But sample throughput for XPS is limited by its ultrahigh vacuum requirements to eight samples per day with a typical four-position sample changer. XRF is much faster and can examine up to 168 samples in an overnight run. The nitrogen analysis takes about 10 min, but it only works well with >1% original blood concentration. FTIR-PAS takes 10–15 min, and FTIR-ATR takes about 3 min.

XRF, he says, proved to be the most suitable technique for following the removal or build-up of protein on cloths. It is very fast and precise. XPS is useful because it can detect protein on washed cloths where the residue is usually located on the outer few nanometers of the fiber. Currently, van Dalen is working on adding dyes that bind to protein stains on a cloth sample so that they can be analyzed by fluorescence or UV-visible spectrometry with diffuse reflectance.

With such techniques having been proven, detergent scientists at Unilever and other companies can take a closer look at those “whiter-than-white” claims in the ads and perhaps come up with a product that truly washes out even the most stubborn of stains. That’s the best news for parents looking at mud-soaked offspring.

Further Reading

1. van Dalen, G. X-Ray Spectrom. 1999, 28, 149–156.
2. van Dalen, G. Appl. Spectrosc. 2000, 54, 1350–1356.

David Bradley is a freelance science writer living in Cambridge, United Kingdom. Send your comments or questions regarding this article to tcaw@acs.org or the Editorial Office 1155 16th st N.W., Washington, DC 20036.