Protocols for Th-234 and Total Mass in the              
               Water Column and Float Traps

    Jennifer Young, Jim Murray, Barbara Paul, Thomas Chapin

Th-234 and Mass water column samples

     We collected seawater from 30-l Go-Flo bottles at typically
8 depths in the upper 250 m per station (e.g., 0, 25, 50, 75,
100, 150, 200, 250m).  Approximately 80 (+/-10) liters were
filtered in parallel through three preweighed 0.4-um, 142-mm
Nuclepore filters.  Particulate activities and total suspended
matter were determined from the pre-weighed filters.  Filters
were rinsed with ~20 ml of deionized, distilled water buffered
with NaHCO3 to pH ~8 to remove salts, folded, and stored in
plastic containers for the return to the laboratory at the
University of Washington.  Total activities were determined in
approximately 20 liters of unfiltered seawater taken at the same
depths as the particulate samples.

     The radioisotopes were isolated and purified using the
methods of Fleer, 1991 and Anderson and Fleer, 1982.  The major
modification is determination of Th-234, Pb-210, Po-210 on the
same sample, rather than two separate samples.  This modification
was developed by Kenneth Coale (unpublished).  Total samples were
acidified with 50 ml of 12 N HCl and approximately 30 dmps of Th-
230 and 36 dmps of U-236 as yield tracers (yield tracers of Po
and Pb were also added), and 150 mg of acidified, resin-cleaned
(1-x8 Biorad) Fe carrier (as FeCl3 or Fe(N03)3) were added.  The
samples were then vigorously shaken and the Th-230 spike was
allowed to equilibrate for at least 6 hours.  Preliminary
experiments with Puget Sound water samples revealed 6 hours was
more than sufficient time for equilibration. 

     In order to scavenge the thorium isotopes, Fe(OH)3
precipitates (pH ~8.5) were formed by addition of 50 ml of
concentrated NH4OH.  The precipitates were allowed to settle for
at least 10 hr.  The supernate was siphoned off and the
precipitates were centrifuged, dissolved in ~10 ml of 12 N HCl,
digested in a Teflon beaker with concentrated HCl, HNO3 and HF,
reprecipitated with NH4OH, and rinsed twice with distilled,
deionized water.  Enough 12 N HCl was added to the precipitates
to make a 9 N solution.  The samples were then placed on an HCl-
conditioned anion-exchange resin (1-x8 Biorad) column.  Th (and
Pb) isotopes were eluted with ~60 ml of 9 N HCl.  U isotopes (and
the Fe) were then eluted with ~30 ml of 0.1 N HCl.  Po isotopes
were eluted with at least 300 ml of 0.01 N HNO3.  The Th and U
fractions were further purified by eluting through two HNO3-
conditioned anion-exchange resin (1-x8 Biorad) columns.  After
purification, the Th fractions were either electroplated onto a
stainless steel planchet (EqPac I) or evaporated to 1 small drop
of HNO3, extracted into a TTA (thenyltrifluoroacetone)-benzene
solution  and then stippled onto a stainless steel planchet
(EqPac II).  Particulate samples were processed similarly, except
the digestion utilized NH4OH, HCl, HNO3, HF and HClO4.
     
     Th-234 (actually daughter nuclide, Pa-234, t1/2 = 6.75 hr)
activities were measured in low-background (0.22-0.75 dpm) gas-
flow beta detectors with lead shielding and anti-coincidence. 
Total and dissolved Th-234 samples were beta-counted at sea,
whereas trap and particulate Th-234 were processed and beta-
counted at the University of Washington.  Beta counter
efficiencies were frequently determined throughout the total
counting period.  The sample plates were wrapped in three layers
of commercial-grade aluminum foil, in order to shield alpha rays
and low level beta rays.  The recoveries of the Th-230 yield
tracers were measured by alpha counting on silicon-surface
barrier counters (EG&G Ortec 576) at the University of
Washington.  The alpha counters were calibrated with an Am-241
standard.  The recovery of Th averaged 40%.  Total counts for the
samples generally exceed 1000 counts, thereby achieving a
statistical standard deviation of < 3%.  The correction for
ingrowth of Th-234 from U-238 during the equilibration time
averaged 4.3% of uncorrected data.  Activities of U-238 were
obtained from the relationship of U-238 (dpm^-1) = 0.0708(+/-
0.0003) x salinity, (Chen et al., 1986)  We assumed the
correction for ingrowth during the time between precipitate
isolation to the first anion-exchange column to be negligible,
since this time averaged 19 hours.

     The Nuclepore filters were dried and re-weighed in order to
determine suspended matter concentrations.  The filters were
soaked in 3-5 ml of deionized distilled water for about an hour
to determine a salt-correction for the sample weight.  The
average salt-correction is 4 mg; the average final correction is
43% of the uncorrected weight.


Protocols for Th-234 and Mass sediment-trap samples

     Modified Moss Landing Marine Laboratories (Knauer et al.,
1979; Martin et al., 1987), surface-tethered, floating sediment-
traps were deployed at 5 depths (between 75 and 250 m) for an
average of 50 (+/-20) hours (except for stations 1 and 2 during
EqPac I where traps were at 7 depths).  Brine (85 ppm Cl) was
placed in the bottom quarter of the traps and overlain by
filtered seawater.  The aspect ratio without brine is 8:1 and to
the top of the brine is 6:1.  Generally three traps at each depth
were used for our Th-234 flux samples.  Mass flux data were
obtained by filtering the contents of each Th-234 trap through a
pre-weighted 0.4 um, 90 mm Nuclepore filter fitted in the bottom
of the trap.  The filters were carefully hand-picked for
swimmers, rinsed with NaHCO3 to pH ~8 to remove salts, then
folded and stored in plastic containers for return to the
laboratory.  Swimmers were kept for weight and Th-234 content
determinations.

     The same procedures for processing Th-234 particulate
samples were used for Th-234 sediment-trap samples and swimmers. 
The average salt-corrections for sample weights is 1 mg per
filter; the average final correction is 20% of the uncorrected
weight.  Because of our relatively short deployment time (~50
hr), we are able to correct for any decay during the deployment
by decay-correcting the measured Th-234 activity in the trap
samples to the mid-point of the deployment and dividing by the
duration of the deployment.

     Blanks for yield tracers, Fe, and unused filters from the
same batch as sample filters were processed for every type of Th-
234 sample.  The blanks were consistently small (< 0.6 dpm).

References Cited

Anderson, R.F. and A.P. Fleer (1982).  Determination of natural
     actinides and plutonium in marine particulate material. 
     Analytical Chemistry, 54, 1142-1147.

Chen, J.H., R.L. Edwards and G.J. Wasserburg (1986).  U-238 and
     Th-232 in seawater.  Earth and Planetary Science Letters,
     80,  241-251.

Fleer, A.P. (1991).  Updated determination of particulate and
     dissolved Th-234.  Marine Particles:  Analysis and
     Characterization.

Knauer, G.A., J.H. Martin and K.W. Bruland (1979).  Fluxes of
     particulate carbon, nitrogen, and phosphorus in the upper
     water column of the northeast Pacific.  Deep-Sea Research,
     26A, 97-108.

Martin, J.H., G.A. Knauer, D.M. Karl and W.W. Broenkow (1987).  
     VERTEX: carbon cycling in the northeast Pacific.  Deep-Sea
     Research, 34, 267-285.