JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 99, NO. B10, PAGES 19,821-19,828, OCTOBER 10, 1994 Fourier transform IR-determination of intragranular water content in quartzites experimentally deformed with and without added water in the ductile deformation field Bas den Brok 1 HPT-laboratory, Institute of EarthSciences, University of Utrecht, the Netherlands J6rg Meinecke and Klaus R611er Institute of Geology, Ruhr University, Bochum, Germany Abstract. To testwhether the weakening effectof added waterin ductile deformation experiments on quartzite at highP-T conditions is associated with measurable changes in intragranular water content (Cw) we determined Fourier transform infrared (FTIR) spectra of single grains (grain size150-250 gm; IR spot size -70 gm) in samples of Dongelberg quartzite experimentally deformed in a Griggs solid-medium deformation apparatus at a temperature of 800øC, a confining pressure of •-1200 MPa, and a strain rateof ~10 -7 s -i, both with and without -0.4 wt % added water. The added water caused a weakening effect of a factor >9. Microstructural observations indicate thatthe dominant deformation mechanism in the weakened samples was solution-precipitation creep. The FTIR measurements showed thatthestarting material has a broad variation in Cw aswell asin the typeof OH absorption. The Cw of individual grains range from <200-2900molarppmH/Si (measurement resolution: 100-1000 molar ppmH/Si; standard error of average Cw values 300-900molar ppmH/Si). The experimentally deformed samples, whether deformed with or without added water,showed a broadly similarvariation in Cw with similar types of OH absorption. Relatively "dry" grains (<100-500 molarppm H/Si) were still present in thesamples deformed with added water, aswell as relatively "wet" grains (>3000molar ppm H/Si) in samples deformed without added water. We were unable to demonstrate any significant changes in average Cw after either of thetreatments. Apparently, thewater weakening effects seen in our experiments occurred without measurable changes in Cw. Introduction It is well known that addition of small amounts of water can strongly affect the flow behavior of natural quartzite in experiments at elevated P-T conditions. For example, samples of natural quartzite (grain size 100 to 2•0 •m), experimentally deformed at 800øC, a confining pressure of-1•00 MPa, with a strain rate of 10-•s 'l, in the presence of-0.4 wt % of added water, are ~10 times weaker than samplesdeformed without addedwater [Jaoul et al., 1984]. Though poorly understood, it is widely assumed that this weakening effect is caused by bulk lattice and/or dislocationpipe diffusion of (a significantpart of) the water into the original quartz grains and subsequent promotion of dislocation mobility (glide and climb) and/or dislocation multiplication [e.g., Paterson, 1989' Tullis and Yund 1989; Kronenberg et al., 1990; Tullis, 1990; Tullis et al., 1990]. Den Brok and Spiers [1991] recently proposeda different explanation for the weakening effect of added water in experimentson natural quartzite. According to these authors, the dominant deformation mechanism during macroscopic ductile deformation of the samples with added water is not dislocation creep, but solution precipitation creep plus microcracking, and they arguedthat theseprocesses cause the tNow at Department of Geology, University of Mainz, Germany. Copyright 1994by the American Geophysical Union. Paper number 94lB01473. 0148-0227/94/941B-01473505.00 observed weakening effect. The reason that den Brok and Spiers [1991] came to this conclusion was that they observed that the dominantdeformationmicrostructure in their samples deformed at a temperature of 800øC, a pressure of- 1200 MPa, a strain rate of 10 '7 s 'l, and with -0.4 wt % added water consisted of abundant small, new polygonal to euhedral quartz grains developed in axially oriented microfractures and grain boundaries. These axial arrays of small new grains contained abundantvoids and channel structures, and it was inferred that the new grains were formed by precipitation from solution. The original grains showed abundantmicroscalesubeuhedral syntaxial overgrowth features on grain boundariesoriented parallel to the shortening direction (Z); such features were conspicuously absent at grain boundaries oriented perpendicular to Z. Two more observations were consistent with solution- precipitation creep plus microcracking which would be difficult to explain if dislocation-plastic mechanisms were assumed to be dominant. First, in samples deformed to -40 and -46% bulk finite strain, no significant lattice preferred orientation was developed [see also den Brok, 1992], and second, samplesdeformed with added water showed a stress- strain rate relationship fitting a power law with a stress exponentn<l.3. Such very low n values were reportedearlier (e.g., n--1.4 by Jaoul et al. [1984]) and were considered enigmatic, since these were difficult to explain when weakeningoccurred by an intracrystalline dislocation-plastic mechanism.They are explained though, when it is assumed that solution-precipitation creep contributedsignificantly to the deformation. 19,821