At least four major quartz-normative and a minimum of two olivine normative magma types have been defined among the continental tholeiites from Massachusetts to North Carolina. Analyses of basaltic lava flows and chilled margins of intrusive diabase sheets and dikes associated with rift-related early Mesozoic sedimentary basins are the basis for the definitions of magma types. The study of these magma types has been evolving over the past two decades through the work of many petrologists (Weigand and Ragland, 1970; Smith and others, 1975; Ragland and Whittington, 1983; Puffer and Philpotts, 1988; Tollo and Gottfried, 1989). In South Carolina and Georgia, south of the area covered by this suite of reports, other olivine-normative magma types have been defined in early Mesozoic dikes (Georgia: Gottfried and Arth, 1985; South Carolina: Warner and others, 1985).

Diabase dikes that strike north, northeast, and northwest cut strata of the early Mesozoic basins and intrude the adjacent Piedmont crystalline and folded Paleozoic terranes. Regardless of strike, all principal magma types are represented, although quartz-normative dikes are more common northward from central Virginia, and olivine-normative dikes are more abundant to the south. Similarly, olivine-normative diabase sheets are exposed in North Carolina, whereas quartz-normative sheets are widespread from Connecticut to central Virginia.

Table 1 summarizes the ranges of selected major elements and some trace element average ratios that have been used to help define the principal magma types. As no single ratio or range in elements is totally effective in defining magma types, major element and several element ratios should be used together (see Table 1). As not all magma types present are documented in this report, we have used the following magma type terminology: HTQ = High Ti, quartz normative; HFQ = High Fe, quartz normative; LTQ = Low Ti, quartz normative; ON = Olivine normative.

From 1984 to 1990, the U.S. Geological Survey carried out a field geochemical and petrologic assessment of early Jurassic diabase intrusives associated with the exposed early Mesozoic basins. Major, trace and platinum group elements were analyzed for more than 960 samples as part of the Survey's Strategic and Critical Minerals Program. Although the results of some of the studies based on these analyses have been published (Froelich and Gottfried, 1988; Gottfried and Froelich, 1988; Gottfried and others, 1990), the complete geochemical data base has never before been provided. Included in this study are powders of diabase samples from local studies by both USGS and non-USGS workers, documented elsewhere, that were analyzed for selected elements by the USGS in the course of these studies. Samples generally labeled FG- in the tables, but not labeled on the map, were collected by Froelich and Gottfried of the USGS in the course of this regional study. Samples collected by others are identified, described, and referenced in the brief text accompanying the map (scale 1:125,000) of each of the eight basin areas documented in this study. A very brief discussion of the local PGE distribution in the diabase is also included with these texts.

Of the principal magma types recognized in the Mesozoic tholeiitic province of the eastern United States, the high Ti quartz-normative (HTQ) variety and some of the olivine normative (ON) types have similar original abundances and ratios of Pd and Pt; about 10 parts per billion (ppb) each in chilled margin samples throughout the province. The other quartz-normative varieties, especially those that formed sheets (LTQ and HFQ), have Pd and Pt abundances commonly less than 5 ppm in the chilled margins and nowhere show significant enrichment. Both ON and HTQ cumulate zones enriched in olivine"or orthopyroxene show slight enrichment of Pt (30 to 50 ppb) and depletion of Pd (less than 5 ppm). In contrast, late stage ferrogabbro, ferrodiorite, and granophyre differentiates are significantly enriched in Pd (100-500 ppb) in at least four of the HTQ sheets and one ON sheet. As the enriched zones are relatively thin (tens of meters), it is likely that many such zones in other thick HTQ sheets were not sampled in the course of this reconnaissance study.

Table 1. Classification of magma types of early Mesozoic diabases and basalts based on characteristic major element abundances and selected trace-element ratios (modified from Tollo and Gottfried, 1989, and published and unpublished U.S. Geological Survey data; quartz-normative data are based on more than 200 analyses).

Magma type         Wt % major elements (Avg basalts and diabases)     Trace element ratios (Avg. basalts)

-------------------------------------------------------------------------------------------------------------------

Quartz normative         Si02             MgO         TiO2         Fe2O3T     100 Nb/Ti     Th/Hf     Hf/Ta     La/Yb(n)

-------------------------------------------------------------------------------------------------------------------

HTQ (High Ti)             50-53         7-8           1.0-1.2         <13                .10            .85          4.7         3.2

LTQ (Low Ti)             49.5-52      7.5-8.5     .7-.85          <13                  .07           .90          6.0         1.7

HFQ (High Fe)            51-54        5.2-6         1.0-1.2        12.5-15           .07            .94         6.6         2.1

HFTQ (High Fe-Ti)     51-54        4.5-6.0     1.3-1.5         >15                 .05            .76           8.0         1.4

-------------------------------------------------------------------------------------------------------------------

Olivine normative (Average dikes and sheets) (Average dikes and sheets)

--------------------------------------------------------------------------------------------------------------------

ONA                         47-49         8-15         .25-.45         <13                 .12             1.2         5.0             1.8

ONB                         47-50         7-24         .35-.8            <13                 .07             .35         7.0            1.35

--------------------------------------------------------------------------------------------------------------------

Fe2O3T = total iron (FeO+Fe2O3). La/Yb(n) = ratio normalized to chondrites. < = less than; > = more than.
 
 

Chemical analyses listed in this report were done in the Reston, Virginia and Denver, Colorado laboratories of the U.S. Geological Survey over a period of eight years (1984 to 1991). All of the chemical procedures that were used are listed below. Because methodology and data requirements varied over time, many elements were determined by more than one analytical procedure. For some cases, ranges of laboratory numbers, corresponding to the numbers in the data tables, are listed next to the name of a method description, indicating which samples were analyzed by which methods. However, data for some elements are primarily taken from a "preferred" method, with missing values filled in with data from secondary methods. In these cases, it is impractical to specify which samples were analyzed by which methods. All intermethod biases are believed to be less than 10%. For elements where ranges of laboratory numbers are not specified, the orders of preference for analytical methods are as follows (abbreviations: flame atomic absorption spectrometry (FAA), graphite. furnace atomic absorption spectrometry (GFAA), inductively coupled plasma atomic emission spectrometry (ICP-AES), instrumental neutron activation analysis (INAA), wavelength dispersive X-ray fluorescence spectrometry (WDXRF), and energy dispersive X-ray fluorescence spectrometry (EDXRF)):

Na20: 1) INAA, 2) "Rapid rock" (see description below) and WDXRF, 3) ICP-AES

CaO: 1) "Rapid rock" and WDXRF, 2) INAA

Cr: 1) INAA, 2) GFAA, 3) ICP-AES

Fe203 (total): 1) INAA, 2) "Rapid rock" and WDXRF, 3) ICP-AES

Co: 1) INAA, 2) FAA, 3) ICP-AES

Ni: 1) ICP-AES, 2) FAA, 3) EDXRF

Cu: 1) EDXRF, 2) FAA

Zn: 1) EDXRF, 2) INAA, 3) ICP-AES

Rb: 1) EDXRF, 2) INAA

Sr: 1) EDXRF, 2) ICP-AES Y: 1) EDXRF, 2) ICP-AES

Ba: 1) EDXRF, 2) INAA, 3) ICP-AES
 
 

Digital versions of the data tabulated in this report are available on double-sided high-density (1.2 MB) 5-1" floppy diskettes (Grossman et al., 1991a) and double-sided high-density (1.44 MB) 3-1" diskettes (Grossman et al., 1991b). These data are in a form compatible with Lotus 1-2-3 Release 2 or its equivalent on IBM-PC-type computers.
 
 

Sampling.

Representative samples of fresh core, quarry, or bedrock outcrops of at least 2 kg were collected, labeled, megascopically identified, and located on 1:24,000 topographic quadrangles by latitude and longitude. Where possible, weathered rinds, veins, and obviously altered material were removed and discarded. In the laboratory, the rock samples were re-examined to assure that all undesireable portions were removed; if any extraneous material was present, it was cut away with a diamond saw and discarded. In general, 2-3 cm-wide, square thin-section blanks were cut from representative specimens.

Grinding.

Large rock fragments (>5 cm) are crushed to -8 mesh (2 to 3 mm) in a jaw crusher with adjustable, hardened Mn-steel plates. The entire sample is run through a "Vertical" grinder with alumina plates, producing a powder of -100 mesh (i.e., most particles are <150 gm). Grinding equipment is ~leaned before each sample by processing quartz-sand.

Splitting.

To prepare aliquants for use in different laboratories, the rock powder is thoroughly mixed, poured onto glassine paper, and divided into four piles. Small portions are taken from throughout each mound with a spatula, and placed in a new bottle. For large samples, cone splitters or riffle splitters are used to obtain aliquants.

"Rapid Rock" method (Lab numbers < W-238427)

Following the method of Shapiro et al. (1975), 200 mg of samples are fused with a lithium meta-borate/tetra-borate flux. The fused sample is dissolved in dilute HNO3. Silicon, Al, P, and Ti are determined colorimetrically. Total Fe, Ca, K, Mg, Mn, and Na are determined by FAA. Fe2O3 is calculated by difference from total Fe and FeO (measured by the method listed below). All data are reported as oxides.

Wavelength-dispersive XRF (Lab numbers > W-238427)

Major element oxides are determined in rocks and minerals by WDXRF (Taggart et al., 1987). 800-mg samples are weighed and then ignited in a Pt-Au crucible at 900 to 9250C for 45 minutes. After cooling, the samples are reweighed to determine the total loss on ignition (LOI). The ignited samples axe fused with 8 g of lithium tetraborate by heating at 11200C for 40 minutes, poured into molds, and the resultant glass disc is irradiated by X-rays generated by a Rh-target tube operating at 35-kV and a current of 60 mA. Characteristic X-rays emitted by each element in the sample are counted, corrected for matrix effects using the dejongh (1973) model, and concentrations are determined using previously prepared calibration standards. Concentration data are then recalculated to account for any mass change on ignition.
 
 

FeO (Peck, 1964).

A 500-mg sample is decomposed using HF and H2SO4. This solution is treated with boric, sulfuric, and phosphoric acids. Fe (II) is determined by a colorimetric or a potentiometric titration with potassium dichromate. Sodium diphenylamine sulfonate is used as the endpoint indicator in the colorimetric titration.

C02 (Engleman et al., 1985).

A 500-mg sample is digested with HC104- C02 is evolved and carried into a coulometric cell. The C02 is converted into a strong acid by ethanolamine, and is titrated coulometrically.

H20- (Shapiro, 1975).

A 1-g sample is weighed and dried at 110'C for a minimum of 1 hr. After cooling in a dessicator, the sample is weighed again and H20- is calculated by difference.

H20+-

Method of Shapiro (1975) A 1-gram sample is fused with 2 grams of anhydrous sodium tungstate. Water in the sample is driven off and collected on a preweighed filter paper. Total water is determined by reweighing the filter paper. H20+ is calculated as the difference of total H20 and previously determined H20-.

Method of Jackson et al.(1987) (Lab numbers W241388-828, W243816-838). A 50mg sample is mixed with 150 mg of lead oxide/lead chromate flux. The sample is heated to 9500C. The evolved water is determined coulometrically by Karl-Fischer titration. This gives the total H20 in the sample. H20+ is determined from the difference between total H20 and H20--

A 20G-mg sample is weighed. Vanadium pentoxide is added as a combustion aid. The sample is combusted in a sulfur analyzer and the sulfur dioxide evolved is measured by an IR detector (Kirschenbaum, 1983).

A 200-mg sample is decomposed using HF, H2SO4, and KMn04 in a specially designed, sealed teflon container. Chlorine is captured in a KOH/Na2SO3 solution in a center compartment of the container. Chlorine is determined as chloride by the selective ion electrode (SIE) method of Aruscavage and Campbell (1983).

Method of Kirschenbaum (1988).

A 100-mg sample is fused with a Na2CO3/ZnO flux. The fusion cake is leached with H20. HCI is added to expel C02. Am aliquot of the sample solution is buffered with a sodium citrate/KN03 solution. This solution is analyzed for fluorine, as fluoride, by the selective ion electrode (SIE) method of Kirschenbaum (1988).

Method of Hopkins (1977) (Lab numbers W241388-828, W243816-838).

A 50-mg sample is fused with a Na2CO3/K2CO3/KNO3 flux. The fusion cake dissolved in 1M citric acid. The sample solution is buffered with sodium citrate. This solution is analyzed for fluorine, as fluoride, by a standard addition method using the SIE method of Hopkins (1977).

Fire Assay/GFAA method.

A 15-gram sample and 2 mg'of gold wire are fused with a PbO/Na2O3/Borax glass/flour flux for .45 min. at 1000*C. The lead button is separated from the fusion slag. Bone-ash cupels are preheated at 1000*C, and the lead buttons are added. After the buttons have melted the temperature is reduced to 850'C, and air is passed over the cupels. The lead is oxidized to PbO, and is absorbed into the bone ash, leaving a gold bead which is dissolved in aqua regia and evaporated to dryness. The residue is taken up in 6M HCI, and analyzed for Pd, Pt and Rh by GFAA (Aruscavage et al., 1984).

The accuracy and precision,of Pd and Pt analyses by this method was monitored by including replicate samples of USGS standard rocks DNC-1, W-2, and BHVO-L (see Gottfried and Froelich, 1988, for a full discussion).

Nickel sulfide/ICP-MS method (Lab numbers W241388-828, W243816-838).

A 10-g sample is.fused with a Ni/S/Si02/Na3CO2/borax-glass flux for 2 hours at 1050'C. The resulting NiS button is removed, broken into pieces, and placed into a test tube. 1 mL of 20% SnC12, 0-1 mL of 1% tellurium solution, and 50-70 mL of concentrated HCI are added to the tube to dissolve the NiS button. The solution is filtered and the residue is collected on polycarbonate filter paper. The filter paper is dissolved in HCI and HN03, diluted to 10 mL with 1% HCI and the solution is analyzed for Pd, Pt, Rh, Ru and Ir by inductively coupled plasma mass spectrometry (ICP-MS).

Approximately 1.0 g of 100-mesh, powdered sample is pressed into a Mylar cup. Samples are analyzed for Ni, Cu, Zn, Rb, ST, Y, Zr and Ba using a Kevex 700 EDXRF spectrometer with a Kevex 7000 analyzer (Johnson, 1984; Johnson and King, 1987). The secondary targets used to fluoresce each element were:

Ni, Cu and Zn : Germanium

Rb, Sr, Y and Zr : Silver

Ba : Gadolinium

Corrections are made for background interferences, escape peaks, and spectral overlaps. Sources of error inherent to EDXRF analysis are corrected using the Compton ratio method. Trace element concentrations in the samples are calculated from calibration graphs of the intensity ratio vs. concentration for a series of standard reference materials found in Abbey (1983).

Multielement Neutron Activation Analysis

Sample aliquants of -0.5 g each are irradiated for ~8 hours at a flux of ~2x 10¹² n-cm-2 -s-l in the "TRIGA" reactor at the U.S. Geological Survey, Denver, CO. Standards for most elements are aliquants of a powdered natural obsidian spiked with primary solutions, taken to dryness and homogenized. Standards for Ca, Ti and Au are powdered CaCO3, TiO2, and homogeneous low-Au quartz, respectively. At least one replicate sample plus USGS standard rocks DNC-1 and W-2 were irradiated together with the samples

and standards.

Samples are counted three times on co-axial Ge and/or Ge(Li) detectors with resolutions ranging from 1.78 to 1.86 KeV measured at 1.33 MeV using the following scheme: 1-hour counts after 6-8 days of decay, 2-hour counts after 14-17 days of decay, and 2-4 hour counts -50 days after irradiation. In addition, one count is done on an intrinsic Ge, low-energy photon detector (for 1 hour, 8-10 days after irradiation).

Gamma-ray spectra are analyzed for Na, Ca, Sc, Cr, Fe, Co, Zn, As, Rb, Zr, Sb, Cs, Ba, La, Ce, Nd, Sm, Eu, Tb, Yb, Lu, Hf, Ta, Au, Th and U. Computer processing is done with SPECTRA and associated programs on a VAX 11/780 computer (Grossman and Baedecker, 1987; Baedecker and McKown, 1987; Baedecker and Grossman, 1989) Corrections are made for spectral interferences as well as for interferences on Zr, Ba, La, Ce and Nd from the products of "'U fission in the reactor. Data for Na, Ca and Fe are reported as oxides.

Denver method (Ba, Cr, Co, Ga, Ni, Sr, Y, and Zn).

Trace elements are measured using the ICP-AES method of Lichte et al. (1987) and Crock et al. (1983). The sample is decomposed using a mixture of HCI, HN03, HC104, and HF at low temperature. The digested sample is taken to dryness, the residue treated with aqua regia, and the solution brought to 10 grams with dilute nitric acid. The solution is nebulized into the ICP-AES plasma discharge.

Reston method (V, Fe, Na)

A 100-mg sample is decomposed with HNO3, HC1O4, and HF and evaporated to dryness overnight. The residue is dissolved in 10 mL of 2N HCI. Analysis of V, Fe and Na is done directly on this solution by ICP-AES (Lichte et al., 1987). Dilutions may be required for higher concentrations of major elements. If the samples are low (<10 ppm) in V the solutions are analyzed directly by GFAA (Aruscavage and Crock, 1987).

Following the method of Aruscavage and Campbell (1979), a 200-mg sample is decomposed using HN03, HC104, and HF, and evaporated to dryness overnight. The residue is dissolved in 15 mL of 20% tartaric acid. The Ag is extracted into butyl acetate as the diphenylthiourea complex. The organic layer is analyzed for Ag by GFAA. Standard solutions are taken through the procedure along with the samples.

A 100-mg sample is decomposed with HNO3, HC1O4, and HF and evaporated to dryness overnight. The residue is dissolved in 15 mL of 1ON HCl. The solution is passed through an ion exchange column to remove the alkali metals. The chloride form of Nb is adsorbed onto the resin. The column is washed with 5N HF to remove Fe. A solution of 7N HNO3 is poured through the column to quantitatively strip the Nb from the resin. This fraction is collected and evaporated to dryness. The residue is dissolved in 2 mL of 2N HCl and analyzed by ICP-AES.

A 100-mg sample is decomposed with HNO.3, HC104, and HF and evaporated to dryness overnight. The residue is dissolved in 10 mL of 2N HCI. Analysis of Co, Cu and Ni is done directly on this solution by FAA (Aruscavage and Crock, 1987). If the samples are low (<10 ppm) in Co, Cr and Cu the solutions are analyzed directly by GFAA (Aruscavage and Crock, 1987). Ni at <20 ppm is analyzed by GFAA using a magnesium nitrate matrix modifier.

10 mg of powdered sample is thoroughly mixed and ground with 40 mg of copper hydroxyfluoride (CUOHF) in an agate mortar. Boron is measured on an emission spectrograph using the method of Golightly et al. (1987), based on the volatilization of boron fluorides into a direct-current arc. The CUOHF inhibits the formation of refractory boron carbide, on the graphite electrode. The concentration of boron, which may range from 0.2 to 600 ppm, is calculated from the intensity of the light emitted from the arc of the sample electrode relative to that of artificial and natural standards run under similar spectrographic conditions.

The accompanying tables were generated by computer from a large data base. Several general rules governed the display of significant figures. (1) Data for major elements are all given in wt.% to two decimal places. (2) Data for trace elements are given to enough figures that the round-off error is3% and averages -1%. These rules rarely lead to the display of too few significant figures, but may result in extra trailing zeros when a laboratory reported fewer digits due to low precision.

Fe2O3* represents total iron calculated as Fe2O3. The totals of major elements (?) have not been corrected for excess oxygen when S, F and Cl are present. Values reported as upper limits were included in the analysis, but were below the detection limit (note that detection limits are matrix dependent by some methods, noteably INAA). Dashes are shown for elements that were not analyzed.