Rare gas systematics on Lucky Strike basalts (37°N, North Atlantic): Evidence for efficient homogenization in a long-lived magma chamber system?



[1] We present rare gas data in fresh glasses from the Lucky Strike segment located on the Mid Atlantic Ridge (∼37.3°N), close to the Azores plateau. We analyzed the helium and neon isotopes in 28 samples by melting as well as He-Ne-Ar-Kr-Xe isotopes in 9 samples by crushing. Samples were collected during the Graviluck06, MOMAR08, and Bathyluck09 cruises over a ridge length of ∼13 km (mean sample spacing of ∼500 m), and at depths ranging from 1550 m to 2174 m. The helium isotopic ratio varies between 84,410 and 88,235 (R/Ra between 8.19 and 8.56). The samples having the “most” primitive helium isotopic ratio are the enriched samples (e.g. high K2O/TiO2) although the difference to the depleted samples is small. It appears that all of our samples derive from the same and well-homogenized magma chamber. Neon isotopes clearly show the influence of the Azores hotspot, which is not seen with helium because of lower 3He/22Ne in the plume source compared to the MORB source.

1. Introduction

[2] Rare gases are excellent tracers of plume-ridge interaction due to the primitive signature seen in Oceanic island basalts (OIB). Indeed, Mid Oceanic Ridge Basalts (MORB) show a 4He/3He ratio of 90,000 ± 10,000 (R/Ra = 8 ± 1 where R is the 3He/4He ratio and Ra is the atmospheric ratio of 1.384 10−6) whereas OIB show lower ratios (higher R/Ra), down to 15,000 (R/Ra = 50) for the ancient Iceland plume [Stuart et al., 2003]. To explain the lower 4He/3He ratios in OIB than in MORB, it is generally proposed that mantle plumes sample a gas-rich reservoir [Kurz et al., 1982]. Neon isotopes show a similar behavior to that of helium isotopes. The 21Ne/22Ne ratio is lower in OIB than in MORB [Honda et al., 1991; Valbracht et al., 1997]. Because the 21Ne is nucleogenic (e.g. 18O(α,n)21Ne or 24Mg(n,α)21Ne), it is generally proposed that in order to preserve a lower 21Ne/22Ne, the OIB source has to be enriched in 22Ne compared to the MORB source. Therefore, both helium and neon are very sensitive to plume-derived magmas due to these expected high concentrations. Previous studies have demonstrated the interest of measuring the rare gases to constrain plume-ridge interactions [Graham et al., 1999; Hopp et al., 2004; Moreira et al., 1995; Poreda et al., 1986; Sarda et al., 2000; Stroncik et al., 2008].

[3] The objective of this study is to constrain the influence of the Azores hotspot along the Mid Atlantic Ridge, as well as to investigate the scale of mantle heterogeneities. We present helium and neon isotopes in 28 samples from the Lucky Strike segment, North Atlantic (37°20′N), as well as He-Ne-Ar-Kr-Xe elemental compositions and isotopic ratios in 9 samples from this segment.

2. Lucky Strike Segment and Sampling

[4] The ∼70 km-long Lucky strike segment is centered at ∼37°20′N on the Mid Atlantic Ridge, and ∼350 km West of the Azores hotspot. It shows a 15–20 km wide axial valley with axial depths that vary from >4000 m at the segment ends to 1550 m at the center of the segment. The volcano located at the center of the segment is ∼10 km wide and 15 km long along the ridge axis, and hosts an active hydrothermal field discovered in 1992 during the FAZAR cruise [Langmuir et al., 1997]. A depression at the summit of the volcano, surrounded by three highs, hosts the hydrothermal fields and a lava lake (Figure 1). An axial magma chamber (AMC) is present at a depth of 3.4 km below the seafloor under the volcano, and extends ∼7 km along axis and ∼3 km perpendicular to it [Singh et al., 2006]. This AMC probably supplies the heat necessary for the hydrothermal field and may be the source of the erupted lava flows. From a geochemical point of view, basalts from the Lucky Strike segment are enriched in incompatible elements [Langmuir et al., 1997]. Samples were collected during the Graviluck06, MOMAR08 and Bathyluck09 cruises over a distance of ∼13 km (mean sample spacing of ∼500 m). Sampling depths vary from 1550 m to 2174 m. Note than we have duplicated a previously analyzed sample from the FAZAR cruise, A127 D15-1 [Moreira and Allègre, 2002] and provided by C. Langmuir. All the analyses are from glass samples, except for two olivines picked from picritic basalts (GRA N16-8 and GRA N17-4) without glass. Two samples (GRARC07 and MOM08V02) are considered as very enriched in trace elements (e.g. high K/Ti ratio) compared to the other samples (A. Bezos, unpublished data, 2011).

Figure 1.

(a) Bathymetric map of the Lucky Strike segment with sample locations. Black dots are N-MORB samples, blue diamonds are picrites (no glass), and red squares are E-MORB according to their higher K/Ti ratio. The insert gives the location of the Lucky Strike segment (LS). Bathymetric data from Cannat et al. [1999], Escartin et al. [2001], Fouquet and Party [1997]. (b) Helium isotopic ratios in Lucky Strike samples reported against latitude. (c) Sample location and projection at the surface of the magma chamber located 3 km beneath the sea floor [Combier, 2007].

3. Analytical Procedure and Results

[5] Samples were cleaned with distillated water and ethanol in an ultrasonic bath. When Mn crust was present, peroxide was used. Two noble gas mass spectrometers were used for this study: ARESIBO II and our new Noblesse (Nu instruments ©). Crushing and melting under vacuum was used for the gas extraction. Only He and Neon were analyzed on the Noblesse whereas the abundances of the five rare gases were determined, as well as isotopic ratios of He-Ne and Ar, on the ARESIBO II mass spectrometer. The analytical procedure on ARESIBO II is given by Moreira and Allègre [2002]. The Noblesse mass spectrometer analytical procedure will be given elsewhere. During the course of the Lucky Strike segment analyses, the precision (1 standard deviation) of the standards was 0.3% for the helium isotopic ratios and 0.2% and 0.5% for 20Ne/22Ne and 21Ne/22Ne, respectively.

[6] Blanks were 1.4 10−10 ccSTP for 4He and 7 10−14 ccSTP for 22Ne for the crushing experiments whereas they were 1.4 10−10 ccSTP for 4He and between 8 10−13 and 3 10−12 ccSTP for 22Ne for the melting procedure. A single heating step at 1400°C was applied on the samples.

[7] Rare gas results are given in Tables 1 and 2. Bulk helium concentrations in glass samples vary from 3 10−6 to 2.2 10−5 ccSTP/g, typical of MORB, whereas crushing experiments give slightly lower helium concentration (3 10−6 to 8.5 10−6 ccSTP/g). Olivine samples give concentration of 5.2 10−9 and 1.1 10−8 ccSTP/g, typical of concentrations in olivines. The 4He/3He ratio varies from 84,409 to 88,235 for the melted glass samples. Crushed glasses give lower helium isotopic ratios, from 78,540 ± 1540 to 87,790 ± 18,30. The two olivine samples gave 4He/3He ratios of 91,350 ± 3350 and 84,310 ± 1280. All the isotopic ratios are typical of MORB. The two most incompatible element enriched samples (GRARC07 and MOM08V02) do not present helium isotopic ratios different from the depleted samples: 84,409 and 85,005, respectively (Figure 1b).

Table 1. He and Ne Results Obtained by Melting at 1400°C and Crushing on the Noblesse Mass Spectrometer. Concentrations are in CCSTP/g. Uncertainties on Concentrations are 5%. Dup Means Duplicate.
SamplesLat (°N)Long (°E)Depth (m)Weight (g)4He (×10−6)4He/3HeσR/Raσ22Ne20Ne/22Neσ21Ne/22Neσ
GRA-N2-2 dup37.3261−32.260720800.0495.45863463108.370.031.08 10−1110.410.150.03910.0017
GRA-N3-237.3327−32.292919460.0354.78865734158.350.044.06 10−1110.030.060.03130.0010
GRA-N4-137.2920−32.280817430.09610.5869473878.310.046.08 10−1212.070.250.04850.0026
GRA-N7-137.2916−32.282417350.0758.37861204188.390.041.15 10−1110.960.110.03880.0015
GRA-N7-1 dup37.2916−32.282417350.0838.41881153228.200.03     
GRA-N16-137.2581−32.312220270.0503.07863033538.370.039.50 10−1110.020.030.03000.0003
GRA-N16-237.2583−32.296119160.0748.76874492938.260.034.98 10−1110.180.030.03160.0003
GRA-N16-437.2632−32.291618470.0939.71865753608.350.038.48 10−1110.060.030.03060.0002
GRA-N16-537.2683−32.289518030.0679.30858053788.420.041.22 10−1111.110.110.03930.0012
GRA-N16-637.2749−32.288017700.0568.58878003018.230.031.77 10−1110.370.060.03470.0007
GRA-N17-137.3360−32.262221280.2776.14881152998.200.032.01 10−1110.410.040.03250.0003
GRA-N22-237.2264−32.306221470.1309.19846003328.540.039.04 10−1110.150.030.03040.0002
GRA-N22-2 dup37.2264−32.306221470.2017.39867472938.330.032.22 10−1110.330.060.03320.0006
GRA-N22-2 dup37.2264−32.306221470.0728.20873692998.270.031.86 10−1110.350.040.03320.0004
GRA-N22-337.2341−32.303421180.0717.39869772988.310.036.08 10−1211.370.180.03950.0019
GRA-N22-537.2527−32.295618330.1415.02853063698.470.044.81 10−10    
GRA-N22-5 dup37.2527−32.295618330.0797.08855242888.450.031.24 10−1010.050.030.03030.0003
GRA-N22-5 dup37.2527−32.295618330.0817.08861202948.390.032.33 10−1110.630.060.03390.0005
GRA-N22-637.2593−32.294618870.1639.70868442978.320.032.09 10−1110.490.040.03450.0005
B09 ROC0737.2914−32.2817 0.1859.56875803018.250.032.79 10−1110.310.030.03320.0003
B09 ROC0837.2914−32.2817 0.0556.42880083158.210.032.33 10−12    
B09 ROC2037.2911−32.2802 0.20011.7861493578.390.034.64 10−1110.330.030.03240.0003
B09 ROC2137.2911−32.2800 0.0739.98863203648.370.042.28 10−1110.480.040.03470.0006
B09 ROC2237.2911−32.2800 0.0678.69882352998.190.035.53 10−1212.090.360.04990.0040
B09 ROC2337.2905−32.2784 0.0477.08865663138.350.03     
B09 ROC2437.2905−32.2784 0.0908.02872752888.280.032.48 10−1110.480.050.03300.0005
A127 D15-137.2905−32.28251706 10.4879273198.220.031.50 10−109.970.030.03010.0003
GRA RC0737.3350−32.251621600.0512.05844093948.560.047.00 10−1211.780.130.04910.0021
MOM08V0237.2874−32.279816290.02221.7850053008.500.034.33 10−1010.340.030.03260.0003
GRA-N16-8 (olivines)37.2829−32.282215500.2055.18 10−99134633497.910.29     
GRA-N17-4 (olivines)37.3116−32.274517790.4741.14 10−88431112798.570.137.60 10−139.910.080.02960.0006
Table 2. Rare Gas Results Obtained by Crushing on the ARESIBOII Mass Spectrometera
SampleLat (°N)Long (°E)Depth (m) 4He (x10−6)22Ne (x10−12)36Ar84Kr130Xe 4He/40Ar*σ 
  • a

    Concentrations are in CCSTP/g. Uncertainties on concentrations are 5%.

GRA-N2-237.3261−32.26072080 10−104.15 10−123.58 10−14 241 
GRA-N4-137.2920−32.28081743step12.913.674.42 10−111.22 10−127.02 10−15 744 
    step21.371.541.14 10−113.23 10−132.12 10−15 623 
    Total4.285.215.56 10−111.55 10−129.14 10−15 704 
GRA-N7-137.2916−32.28241735 3.635.381.15 10−103.84 10−122.00 10−14 583 
GRA-N16-337.2604−32.29321895 3.524.243.73 10−111.08 10−126.05 10−15 663 
GRA-N16-637.2749−32.28801770 3.2014.23.16 10−109.62 10−125.67 10−14 543 
GRA-N22-237.2264−32.30622147 3.163.442.97 10−118.99 10−135.35 10−15 19110 
GRA-N22-337.2341−32.30342118 3.773.662.95 10−111.06 10−125.46 10−15 20011 
B09 ROC2037.2912−32.2802  5.254.472.85 10−118.96 10−135.26 10−15 764 
B09 ROC2137.2905−32.2800  4.814.002.09 10−116.35 10−135.32 10−15 734 
Sample 4He/3HeσR/Raσ20Ne/22Neσ21Ne/22Neσ38Ar/36Arσ40Ar/36Arσ
GRA-N2-2 8206016178.810.1711.480.130.04190.00080.18880.000413568
GRA-N7-1 8779418328.230.1711.190.140.04010.00090.19020.00038405
GRA-N16-3 8109416388.910.1811.630.180.04560.00130.19000.0005173314
GRA-N16-6 8401716618.600.1710.420.090.03350.00030.18780.00034823
GRA-N22-2 7853715379.200.1811.600.180.04210.00100.18890.00058546
GRA-N22-3 8353118098.650.1911.450.220.03980.00140.19020.000693512
B09 ROC20 8500517338.500.1711.920.140.04560.00100.19000.0005273319
B09 ROC21 8726418408.280.1712.290.160.04810.00090.19110.0005343931

[8] 22Ne concentrations vary from 2.3 10−12 to 4.8 10−10 ccSTP/g for melted samples and from 1.54 10−12 to 1.42 10−11 ccSTP/g for crushing. 20Ne/22Ne ratios vary from 9.97 to 12.09 for melting experiments and from 10.42 to 12.29 for the crushing ones. The 21Ne/22Ne is correlated with 20Ne/22Ne and varies from 0.0300 to 0.0499 (melting) and from 0.0335 to 0.0481 (crushing). Results are plotted on Figure 2. Samples from Lucky Strike fall on a different line than the classical MORB line defined by [Sarda et al., 1988]. Note that both mass spectrometers give the same neon systematics.

Figure 2.

Three-isotope neon diagram. The MORB composition is derived from Sarda et al. [1988]. MORB from the Lucky strike segment show a different slope in this diagram, which can be attributed to mixing with a component having primitive neon. Uncertainties are 1σ.

[9] 36Ar concentrations vary from 1.1 10−11 to 3.2 10−10 ccSTP/g (only crushing) and is correlated to 22Ne abundances (not shown). 38Ar/36Ar ratios are between 0.1878 and 0.1911 (air = 0.1880). 40Ar/36Ar ratios are between 482 and 3439. The highest ratio (3439) is far from the mantle value estimated at ∼30,000 [Staudacher et al., 1989]. 84Kr and 130Xe abundances are correlated with neon and argon abundances. 84Kr is comprised between 3.2 10−13 and 9.6 10−12 ccSTP/g. 130Xe varies from 2.1 10−15 and 5.7 10−14 ccSTP/g. Heavy rare gases will not be discussed in this paper. Indeed, 40Ar/36Ar ratios correlate with the 20Ne/22Ne (not shown). This correlation can be interpreted as the result of a mixture between a magmatic component and an atmospheric component. The [22Ne/36Ar]Magmatic/[22Ne/36Ar]air that is necessary to explain the Ne-Ar correlation is ∼50, far above the ratio observed for undegassed samples such as the popping rock 2πD43, which presents a ratio of 1.6 [Moreira et al., 1998]. Such a strong hyperbolic mixture doesn't allow the determination of the uncontaminated 40Ar/36Ar ratio of the Lucky Strike magma. This high [22Ne/36Ar]Magmatic/[22Ne/36Ar]air ratio clearly reflects the high degassing rate of the samples as illustrated by the high 4He/40Ar* ratios (Table 2).

4. Discussion

4.1. Homogeneity of the Helium Isotopic Ratios and the Evidence for a Well-Mixed Magma Chamber

[10] Figure 1b shows the helium isotopic ratio obtained by melting reported versus the latitude. Ratios show very homogeneous results, with a mean ratio of 86,649 ± 1010 (R/Ra = 8.34 ± 0.09). Olivine samples from two picrites give the same helium isotopic ratio as the glass samples within uncertainties. The isotopic variation (∼1%) is small compared to that reported for a complete ridge (few percents) [Allègre et al., 1995] or a ridge section (∼3% for SWIR [Georgen et al., 2003]).

[11] At that point, it is clear that we are not able to distinguish with helium any peculiar chemical heterogeneity in the mantle source of Lucky Strike segment. If the mantle has a marble cake structure [Allègre and Turcotte, 1986], the enriched component (e.g. pyroxenite) is easily homogenized with the depleted one for helium at the scale of our sampling interval (13 km along-axis). Homogenization could occur before melting by diffusion of helium in the mantle [Hart et al., 2008], homogenization by small-scale convection in the uppermost upper mantle [Graham et al., 2001] or, more probably, after melting in a magma chamber considering the scale of our study. In this last case, we could conclude that all the MORBs from the Lucky Strike segment sample the same magma chamber, which was seismically imaged [Singh et al., 2006].

4.2. Evidence for the Azores Influence on the Lucky Strike Volcano Source

[12] Azores island basalts are characterized by both relatively low 4He/3He ratios (high R/Ra) for the islands from the central group and by radiogenic helium for São Miguel island [Moreira et al., 1999]. The interpretation for these helium isotopic ratios is the presence of a mantle plume deriving from a primitive mantle and centered under the central group, in agreement with melt fluxes [Bourdon et al., 2005] and seismology [Yang et al., 2006], and with the presence of a chemical heterogeneity with crustal affinity under São Miguel island [Beier et al., 2008]. Neon systematics also suggest a primitive signature for the Azores hotspot [Madureira et al., 2005].

[13] The helium isotopic ratios of samples from Lucky Strike are not different from the normal MORB value. Indeed, the mean north Atlantic MORB ratio is 89,380 ± 9300 [Allègre et al., 1995] whereas the mean helium ratio of Lucky Strike samples is 86,480 ± 1180. Therefore, based only on the helium isotopes, the Azores hotspot influence cannot be seen in the Lucky Strike segment.

[14] Neon isotopes give a different view of the Azores influence from that of helium. Indeed, as illustrated by Figure 2, the slope of the Lucky Strike samples in the three-isotope neon diagram is higher than the MORB line defined by [Sarda et al., 1988]. This line is interpreted as reflecting a mixing between a mantle-derived component and an atmosphere-derived component (e.g. crust assimilation, seawater interaction, air introduction in the samples). That different slope than MORB means that the mantle beneath Lucky strike has a lower 21Ne/22Ne than the normal Atlantic mantle and this can be attributed to the Azores hotspot influence. Figure 3 shows the 4He/3He versus the 21Ne/22Ne isotopic ratio extrapolated to 20Ne/22Ne = 12.6 (e.g. corrected for air contamination), which corresponds to the value of the “Neon B”, representative of the mantle value [Raquin and Moreira, 2009; Trieloff et al., 2000]. Mantle plume magmas have lower 3He/22Ne ratio than the MORB magmas (by a factor 5 to 10) [Moreira et al., 2001; Yokochi and Marty, 2004] due either to different 3He/22Ne ratios in the two sources, or to mixing of degassed magmas [e.g., Moreira et al., 2001]. Therefore, in Figure 3, mixing is not linear but hyperbolic, which means that neon is more sensitive to the Azores mantle source influence than helium. The isotopic ratios for Terceira Island (Azores) and Lucky strike segment are reported on Figure 3. One can clearly see that the Lucky Strike magma are a mixture between magma from the normal MORB mantle and magma from a primitive component, being the Azores mantle plume.

Figure 3.

Extrapolated 21Ne/22Ne ratios of the regression line of all Lucky Strike samples (at a 20Ne/22Ne of 12.6) plotted versus the mean 4He/3He ratio. The hyperbola is derived from [Kurz et al., 2009]. The Terceira Island (Azores) isotopic ratios are given by Madureira et al. [2005]. The r parameter is 3He/22NeMORB/3He/22NePrimitive mantle. For most OIB and on-ridge hotspots, r is close to 10 [Kurz et al., 2009; Moreira and Allègre, 1998].

5. Conclusions

[15] Rare gases in MORB from the central part of the Lucky Strike segment show that the mantle beneath this segment is extremely homogeneous. Neon isotopes clearly suggest the influence of the Azores hotspot, although this is not seen with helium isotopes. We conclude that all the analyzed samples derive from the same central magma chamber, and that it is very well homogenized. This is easily achieved for helium, and may be neon, due to their high diffusivity in melt.


[16] We acknowledge the effort and support of the officers, crew, and science parties of the Graviluck'06, MoMAR'08, and Bathyluck'09 cruises and the team operating the Nautile submersible and the VICTOR remotely operated vehicle. Ship time was financed by CNRS/INSU and IFREMER, with additional support from ANR (MOTHSEIM Project NT05-3_42213) to J. Escartin. Philippe Sarda and Pete Burnard are thanked for their very constructive reviews. The editor thanks Pete Burnard and Philippe Sarda for their review.