U-series dating of constructional cold-water corals is a powerful tool to reconstruct the evolution of corals on carbonate mounds. Here we have investigated the time framework of corals such as Lophelia pertusa and Madrepora oculata on five different mound settings of the eastern North Atlantic (on Rockall Bank and in Porcupine Seabight), sampled at variable depth and location (610-880 m water depth). We have found that the past 11 ka reflect a period generally favourable for coral development. We further determined local mound growth rates and identified mound surface erosion (framework collapse) during times of active coral framework construction. "Local" vertical mound growth rates vary between less than 5 cm ka − 1 and up to 220 cm ka − 1 . We interpret rates exceeding 15 cm ka − 1 as representative of densely populated coral reefs. During times of reduced or absent coral development, mound evolution rates are by far smaller (0 to < 5 cm ka − 1 ). The time resolution achieved here furthermore provides first evidence for reduced coral (ecosystem) activity at 1.8-2.0 ka, 4.2-4.8 ka and between 6 and 8.2 ka within the Holocene that may be related to climate driven changes of the coral growth environments. During Glacial periods coral growth in those areas seems apparently extremely reduced or is even absent on mounds.

IR-femtosecond pulses were used at high repetition rates (up to 10 kHz) to ablate viscous crude oils for the determination of trace elements by ICPMS. A special internal glass cap was fitted into the ablation cell to minimise oil splashes and remove big particles that would be otherwise spread into the cell. Laser ablation in static and dynamic conditions (i.e. the laser beam being moved rapidly at the surface of the sample) was studied together with some fundamental parameters like repetition rate and fluence. Signal sensitivity and stability were found to be strongly affected by repetition rate and fluence, though not in linear manner, and in some circumstances by the laser beam velocity. Sample transport efficiency was found to decrease with increasing repetition rate, probably due to stronger particle agglomeration when increasing the density of primary particles. ICPMS plasma atomisation/ionisation efficiency was also found to be affected to some extent at the highest repetition rates. Moderate repetition rate (1 kHz), high fluence (24 J cm(-2)) and fast scanning velocity (100 mm s(-1)) were preferred taking into account signal intensity and stability. Sample transport elemental fractionation was also evidenced, particularly as regards to carbon due to volatilisation of volatile organic species. Matrix effect occurring when comparing the ablation of transparent (base oil) and opaque (crude oil) samples could not be completely suppressed by the use of IR femtosecond pulses, requiring a matrix matching or a standard addition calibration approach. This approach provided good accuracy and very low detection limits in the crude oil, in the range of ng g(-1).