The first accelerated weathering study began on July 24, 2006. This study was conducted by Georgette Lang, a chemistry major at Centenary College of Shreveport, Louisiana under the supervision of Jason Church. For this study two types of marble were cored and prepared. Three replicate samples of marble were prepared for each type of cleaner. Along with these samples, three untreated samples of each marble type were readied as internal standards. Finally, three samples of each marble were prepared that would not be treated in the Weatherometer but remained untreated as control samples. This brought the total number to 48 samples. Each sample was given a unique number which encoded information about marble type, chemical cleaner used and the sample identification number. This unique 3 digit number was inscribed on the back of each sample.
Pre-existing conditions of the marble surface were recorded by using the Laser Profilometer (see section 126.96.36.199.1.) to map the surface of each sample prior to treatment. Each sample was photographed and color measurements were taken to check for any color change as a result of the application of the cleaners. The weight of each sample was recorded as a baseline to identify residual material deposited from the cleaning. Once the samples were documented and mounted into the Weatherometer sample holders, the 800 hour test was initiated.
The samples were sprayed with the select cleaner and rotated inside the Weatherometer on a daily bases. The marble was treated with the six cleaners D2, Daybreak, Kodak Photo-flo, H2Orange2, Marble Cleaner, and water (plus one set that was weathered but untreated). Each of the chemicals was mixed to the manufactures recommendation and applied to the sample using a 16 oz hand pump spray bottle. The samples were removed from the Weatherometer in their holder and sprayed to completely wet the surface at the end of a dark cycle at approximately the same time each day. The end of a dark cycle was chosen as a time for treatment so that the cleaner would have sufficient time to soak into the stone without evaporating at elevated temperatures during the UV exposure. After the sample was sprayed, it was placed back into the Weatherometer without being rinsed. The decision was made not to rinse the marble after it had been treated because 4 out of the 5 cemeteries involved in the study stated that they do not rinse their stones post cleaning.
On August 27, 2006 the Weatherometer run ended and the samples were left in the powered down Weatherometer for 48 hours to allow any moisture in the stone to evaporate. Testing began after the samples were removed from their holders. Once accelerated weathering was concluded, testing repeated using the same methods as pretesting documentation. First, the weight measurement is taken. Second, samples were photographed, and third, colorimetry measurements were taken. Finally, surface texture on each sample was measured using the laser profilometry. Also, at this time each of the elemental composition of sample surfaces were analyzed using the Tracer III portable XRay Fluorescence Spectrometer. The XRF Spectrometer under the following conditions: to the Rhodium target Xray tube was set to 15kv and 15ma. A vacuum pack was connected to the Spectrometer and a vacuum of 2 torr was pulled. All spectra were collected for 180 seconds. Spectra were taken of the front and the back surface of samples from both marble types treated with each cleaner. This helped to determine if any chemical residue had migrated through the sample.
Figure 12. Jason Church uses the XRF spectrometer to analysis a marble sample after artificial weathering.
There were a variety of results from the first accelerated weathering test. Colorado Yule marble was more likely than Cherokee White marble to display deterioration or discoloration in the accelerated weathering test. NCPTT is currently looking into the possible reasons that the Colorado Yule marble samples were affected at a greater rate than the Cherokee White marble samples.
There was no discernable change in the samples that were untreated or treated with water. The Colorado Yule marble samples treated with D/2 discolored and took on a slightly translucent appearance. The backs of the Colorado Yule samples also had a very fine powdery deposit on them. When the backside of one sample was examined with the XRF there was a slight Potassium peak. The Colorado Yule samples that were treated with Daybreak discolored to a yellow appearance and had a fine ìsandyî coating on the backside. When the backside of one Colorado sample was examined with the XRF a large Chloride peak was detected. The remaining 3 cleaners (Photo-Flo, WEG Marble Cleaner and H2Orange2) had no obvious detectable deterioration.
The thirty three day study represents a worst case scenario where the marble was saturated with a cleaner on a regular base and was not rinsed after cleaning. Any physical change to the marble or chemical deposition on the marble would likely be scene in an extreme situation. Because of the severity represented in the first accelerated weathering study the decision was made to start a second accelerated weathering study.
188.8.131.52. Four Time Cleaning Study
The second accelerated cleaning study used the Q- Panel Lab Products model QUV/ Spray Accelerated Weather Tester preformed at NCPTT began on December 22, 2006 and ended its 800 hour cycle on January 19, 2007. The second accelerated weathering test was an abbreviated version of the first experiment. For the second study only Colorado Yule marble was selected for testing. This decision was made due to the fact that deterioration was evident on the Colorado marble in the first experiment. These samples were prepared from the same marble using the same procedure as in the first accelerated weathering experiment.
For the second experiment only 10 samples were prepared for Weatherometer exposure. These consisted of 2 Colorado Marble replicates for each cleaner. Each sample was sprayed with D/2, Daybreak, WEG Marble Cleaner, or water. Two cleaners, H2Orange Cleaner and Kodak photo-Flo, were removed excluded from the accelerated weathering, based on results of phase one of the study. Two untreated samples were weathered in this experiment as internal controls. Each of these samples was given a unique 3 digit number that was inscribed on the back of the stone.
One key feature of this study was the decision to only clean the samples weekly, once at the beginning and then three more times at the same cycle on each following Friday. This cleaning schedule may have provided a more realistic approach to the accelerated weathering. Also in this study, the cleaner was sprayed onto the sample then rinsed shortly after being treated according to manufacturersí suggested cleaning directions.
After the 800 hours of weathering was completed, the samples were analyzed in the same steps as the first experiment including laser profilometry and colorimetry. A few additional tests were added to this study to try and get a more detailed view of the stones reaction to the cleaner. Salt deposits were visible on both the back side of some samples and on the Teflon Weatherometer holder ring that surrounds the stone in place. Gravimetric measurements were taken of each of the stone samples while they were still in the holder. Crystalline grow was visible on the backside of the samples treated with both D/2 and Daybreak.
In addition to the usual photo documentation the samples were also photographed under magnification using both a Leica MZ8 boom microscope at a magnification range of 10x to 50x and a Leica DMRX polarized light microscope at a magnification range of 100x to 500x. Both microscopes were fitted with a Diagnostic Instruments Inc. Digital Spot Camera. Through this process the shape, appearance, growth pattern and relative size of the crystalline growth can be documented.
Figure 13. Examples of salt formation on marble samples treated with D/2 (left) and Daybreak (right) both viewed under 100x magnification. Future research activities will investigate the chemical composition of the efflorescence found on the samples. Both the face and back of all of the samples will be analyzed using the portable XRF Spectrometer. Samples of the salt will be analyzed using X-ray Diffractometry. This will help to determine the bases of the visible salts and any other type of detectable chemical residue left on the stone after cleaning. Treated and weathered Colorado Yule samples are still being tested and the data processed at this time. New testing methods are currently being considered to maximize the identification of the salt contents in the samples and to identify any other chemical or physical changes that may have taken place in the samples as a result of the cleanerís application.