Neutron-capture element abundances in the planetary nebula NGC 5315 from deep optical and near-infrared spectrophotometry

TL;DR

This study investigates neutron-capture element abundances in planetary nebula NGC 5315 using high-resolution optical and near-infrared spectroscopy to understand the s-process operation in its progenitor star.

Contribution

It provides detailed abundance measurements of n-capture elements in NGC 5315 and explores possible evolutionary scenarios of its progenitor star based on the lack of s-process enrichment.

Findings

Detected over 700 emission lines including n-capture ions

Total abundances align with previous studies, showing no s-process enrichment

Progenitor likely in a binary system or a low-mass star with limited s-process activity

Abstract

We analyze the chemical composition of the planetary nebula (PN) NGC 5315, through high-resolution (R = 40000) optical spectroscopy with UVES at the Very Large Telescope, and medium-resolution (R = 4800) near-infrared spectroscopy with FIRE at Magellan Baade Telescope, covering a wide spectral range from 0.31 to 2.50 micron. The main aim of this work is to investigate neutron (n)-capture element abundances to study the operation of the slow n-capture ("s-process") in the AGB progenitor of NGC 5315. We detect more than 700 emission lines, including ions of the n-capture elements Se, Kr, Xe, and possibly Br. We compute physical conditions from a large number of diagnostic line ratios, and derive ionic abundances for species with available atomic data. The total abundances are computed using recent ionization correction factors (ICFs) or by summing ionic abundances. Total abundances of common elements are in good agreement with previous work on this object. Based on our abundance analysis of NGC 5315, including the lack of s-process enrichment, we speculate that the most probable scenario is that the progenitor star is in a binary system as hinted at by radial velocity studies, and interactions with its companion truncated the AGB before s-process enrichment could occur. However there are other two possible scenarios for its evolution, that cannot be ruled out: i) the progenitor is a low-mass single star that did not undergo third dredge-up; ii) the progenitor star of NGC 5315 had an initial mass of 4--6 solar masses, and any s-process enhancements were heavily diluted by the massive envelope during the AGB phase.

Tables

Table 1. Table 1: Journal of observations.

Telescope	Date	$\Delta \lambda$ (Å)	Exp. time (s)	Seeing (′′)
8.2 m VLT	2014/01/02	B1: 3100$-$3885	60, 2$\times$630	1.1
”	”	B2: 3750$-$4995	60, 2$\times$630	1.1
”	”	R1: 4785$-$6805	60, 2$\times$630	1.1
”	”	R2: 6700$-$10420	60, 2$\times$630	1.1
”	2014/02/03	B1: 3100$-$3885	6$\times$630	1.5
”	”	B2: 3750$-$4995	6$\times$630	1.5
”	”	R1: 4785$-$6805	6$\times$630	1.5
”	”	R2: 6700$-$10420	6$\times$630	1.5
”	2014/02/04	B1: 3100$-$3885	4$\times$630	1.4
”	”	B2: 3750$-$4995	4$\times$630	1.4
”	”	R1: 4785$-$6805	4$\times$630	1.4
”	”	R2: 6700$-$10420	4$\times$630	1.4
6.5 m MBT	2013/08/13	8000$-$25000	20$\times$30	1.8–2.9

Table 2. Table 4: Comparison of observed line ratios and those predicted by the Cloudy model.

Ion	Line ratio	Model	Obs.
He i	6934/7161	0.437	0.448
He i	7161/7816	0.308	0.263
He i	10999/12988	0.426	0.492
[N ii]	5755/6538	0.0918	0.0816
[O ii]	3726/3729	2.482	2.494
[O iii]	4363/5007	0.0048	0.0045
[S ii]	6731/6716	2.183	2.014
[S iii]	6312/9069	0.0619	0.0571
[Cl iii]	5538/5518	2.823	2.900
[Ar iii]	5192/7136	0.0057	0.0043
[Ar iv]	4740/4711	3.200	3.398

Table 3. Table 5: Corrected line ratios ( I (H β 𝛽 \beta ) = 100) for n -capture elements in NGC 5315.

$\lambda$0 ($\AA$)	Ion	I($\lambda$)/I(H$\beta$)	Error(%)a
UVES
5346.02	[Kr iv]	0.0081	20
5867.74	[Kr iv]	0.0188	12
6555.56	[Br iii]	0.0097	:
6826.70	[Kr iii]	0.0161	16
7535.40	[Xe iv]	0.0032	:
8854.00	[Se iii]	0.0055	:
FIRE
8855.28	[Se iii]	0.0062	:
10992.00	[Se iii]	0.0120	:
2.1986 $\mu$m	[Kr iii]	0.0556	13
2.2864 $\mu$m	[Se iv]	0.0827	11
a Colons indicates errors larger than 40 $\%$

Table 4. Table 6: Atomic data set used for collisionally excited lines.

	Transition	Collisional
Ion	Probabilities	Strengths
N+	Froese Fischer & Tachiev (2004)	Tayal (2011)
O+	Froese Fischer & Tachiev (2004)	Kisielius et al. (2009)
O2+	Froese Fischer & Tachiev (2004)	Storey et al. (2014)
	Storey & Zeippen (2000)
Ne2+	Galavís et al. (1997)	McLaughlin & Bell (2000)
S+	Podobedova et al. (2009)	Tayal & Zatsarinny (2010)
S2+	Podobedova et al. (2009)	Tayal & Gupta (1999)
Cl+	Mendoza & Zeippen (1983)	Tayal (2004)
Cl2+	Mendoza (1983)	Butler & Zeippen (1989)
Cl3+	Kaufman & Sugar (1986)	Galavís et al. (1995)
	Mendoza & Zeippen (1982a)
	Ellis & Martinson (1984)
Ar2+	Mendoza (1983)	Galavís et al. (1995)
	Kaufman & Sugar (1986)
Ar3+	Mendoza & Zeippen (1982b)	Ramsbottom & Bell (1997)
Fe2+	Quinet (1996)	Zhang (1996)
	Johansson et al. (2000)
Se2+	Sterling et al. (2017)	Sterling et al. (2017)
Se3+	Biémont & Hansen (1986b)	Schoning (1997)
Kr2+	Biémont & Hansen (1986b)	Schoning (1997)
Kr3+	Biémont & Hansen (1986a)	Schoning (1997)
Br2+	Biémont & Hansen (1986a)	Schoning (1997)a
Xe3+	Biémont et al. (1995)	Schoening & Butler (1998)

Table 5. Table 7: Physical conditions in NGC 5315.

Diagnostic
ne (cm-3)
[O ii] $\lambda$3726/$\lambda$3729		30225${}_{-7900}{}^{}{}_{}{}^{+25500}$
[S ii] $\lambda$6731/$\lambda$6716		9400:
[Fe ii]a		30000$\pm$7000
[Cl iii] $\lambda$5538/$\lambda$5518		26800${}_{-6500}{}^{}{}_{}{}^{+9000}$
ne(low-mid) (adopted)		26800${}_{-6500}{}^{}{}_{}{}^{+9000}$
[Ar iv] $\lambda$4740/$\lambda$4711		38150${}_{-9850}{}^{}{}_{}{}^{+12500}$
ne(high) (adopted)		38150${}_{-9850}{}^{}{}_{}{}^{+12500}$
Te (K)
[N ii] $\lambda$5755/$\lambda$6584		9850$\pm$625
Te(low) (adopted)		9850$\pm$625
[O iii] $\lambda$4363/$\lambda$$\lambda$5007		8700$\pm$200
[Ar iii] $\lambda$5192/$\lambda$7136		8350$\pm$350
[S iii] $\lambda$6312/$\lambda$9069		8975$\pm$350
Te(mid-high) (adopted)		8650$\pm$200
a Average value between different diagnostics (see text).

Table 6. Table 8: Ionic abundances from CELs in units of 12 + log(X i + limit-from i {}^{\emph{i}+} /H + ).

Ion	Line used	Abu.UVES		Abu.FIRE
N+	$\lambda$6548	7.67$\pm$0.08
N+	$\lambda$6584	7.70$\pm$0.08
N+	(adopted)		7.69$\pm$0.08
O+	$\lambda$3726	7.96${}_{-0.15}{}^{}{}_{}{}^{+0.17}$
O+	$\lambda$3729	7.96${}_{-0.15}{}^{}{}_{}{}^{+0.17}$
O+	(adopted)		7.96$\pm$0.17
O2+	$\lambda$4959	8.70 $\pm$0.05
O2+	$\lambda$5007	8.72 $\pm$0.05
O2+	(adopted)		8.71 $\pm$0.05
Ne2+	$\lambda$3868	8.11 $\pm$0.07
Ne2+	$\lambda$3968	8.11 $\pm$0.07
Ne2+	(adopted)		8.11 $\pm$0.07
P+	$\lambda$7876	4.10 $\pm$0.12
P+	$\lambda$11470			4.09 $\pm$0.06
P+	$\lambda$11886			4.28 $\pm$0.06
P+	(adopted)		4.24 $\pm$0.09
S+	$\lambda$6716	6.39${}_{-0.10}{}^{}{}_{}{}^{+0.12}$
S+	$\lambda$6731	6.35${}_{-0.10}{}^{}{}_{}{}^{+0.12}$
S+	(adopted)		6.36$\pm$0.12
S2+	$\lambda$9069			7.08$\pm$0.03
S2+	$\lambda$9530			7.13$\pm$0.03
S2+	(adopted)		7.11 $\pm$0.03
Cl+	$\lambda$9123	4.33 $\pm$0.07	4.35$\pm$0.07	4.37 $\pm$0.07
Cl2+	$\lambda$3353	5.16 ${}_{-0.11}{}^{}{}_{}{}^{+0.13}$
Cl2+	$\lambda$5517	5.29 $\pm$0.08
Cl2+	$\lambda$5538	5.29 $\pm$0.05
Cl2+	$\lambda$8434	5.20 ${}_{-0.09}{}^{}{}_{}{}^{+0.11}$
Cl2+	$\lambda$8484	5.31 ${}_{-0.09}{}^{}{}_{}{}^{+0.11}$		5.42 ${}_{-0.10}{}^{}{}_{}{}^{+0.12}$
Cl2+	(adopted)		5.29$\pm$0.07
Cl3+	$\lambda$8046	3.86 $\pm$0.06
Ar2+	$\lambda$7135	6.61 $\pm$0.05
Ar2+	$\lambda$7751	6.62 $\pm$0.05
Ar2+	(adopted)		6.61 $\pm$0.05
Ar3+	$\lambda$4711	4.92 $\pm$0.10
Ar3+	$\lambda$4740	4.92 $\pm$0.06
Ar3+	(adopted)		4.92 $\pm$0.08
Fe+	$\lambda$7154	5.05${}_{-0.08}{}^{}{}_{}{}^{+0.12}$
Fe+	$\lambda$8619			5.12 $\pm$0.07
Fe+	1.2570 $\mu$m			5.17 $\pm$0.06
Fe+	1.2946 $\mu$m			5.18 $\pm$0.07
Fe+	1.3209 $\mu$m			5.16 $\pm$0.07
Fe+	1.3281 $\mu$m			5.20 $\pm$0.09
Fe+	1.5339 $\mu$m			5.21 $\pm$0.07
Fe+	1.5999 $\mu$m			5.15 $\pm$0.09
Fe+	1.6642 $\mu$m			5.35 $\pm$0.11
Fe+	(adopted)		5.15 $\pm$0.08
Fe2+	$\lambda$4701	5.01 $\pm$0.10
Fe2+	$\lambda$4881	5.09 $\pm$0.10
Fe2+	$\lambda$5270	4.96 $\pm$0.11
Fe2+	2.2187 $\mu$m			5.24 $\pm$0.15
Fe2+	(adopted)		5.05 $\pm$0.12
Se2+	$\lambda$8856	3.11:		3.16:
Se2+	1.0995 $\mu$m			3.33 $\pm$0.16
Se2+	(adopted)		3.33 $\pm$0.16
Se3+	2.2864 $\mu$m			2.90 $\pm$0.05
Br2+	$\lambda$6556	2.97:
Kr2+	$\lambda$6826	3.34 $\pm$0.07
Kr2+	2.1986 $\mu$m			3.39 $\pm$0.06
Kr2+	(adopted)		3.38 $\pm$0.06
Kr3+	$\lambda$5346	3.00 $\pm$0.09
Kr3+	$\lambda$5867	3.12 $\pm$0.06
Kr3+	(adopted)		3.08 $\pm$0.08
Rb3+	$\lambda$5759	2.80
Xe3+	$\lambda$7535	2.38:

Table 7. Table 9: Comparison of ionic abundance ratios. The lines used to compute ionic abundances are listed in Table  8 .

Ratio	UVES	FIRE	Peimbert et al. (2004)a
O+/O2+	0.204$\pm$0.088	–	0.153
Cl3+/Cl2+	0.037$\pm$0.009	–	0.036
Cl+/Cl2+	0.110$\pm$0.028	0.089$\pm$0.032	0.089
Ar3+/Ar2+	0.020$\pm$0.005	–	0.011
Fe+/Fe2+	1.041$\pm$0.442	0.844$\pm$0.388	–
a Computed with atomic data shown in Table 6

Table 8. Table 10: Atomic data set used for recombination lines.

Ion	Recombination Coefficients
H+	Storey & Hummer (1995)
He+	Porter et al. (2012, 2013)
C2+	Davey et al. (2000)
N2+	Fang et al. (2011, 2013)
O2+	Storey (1994)
	Liu et al. (1995)
Ne2+	Kisielius et al. (1998)
	Kisielius & Storey (private communication)
Mg2+	Kisielius et al. (1998)

Table 9. Table 11: Ionic abundances from ORLs.

Mult.	$\lambda$0 ($\AA$)	Xi+/H+(10-5)
	He+
14	4471	12267$\pm$860
11	5876	13150$\pm$924
46	6678	12219$\pm$853
	adopted	12819$\pm$880
	C2+
6	4267.15	66$\pm$5
16.04	6151.43	61$\pm$6
17.02a	9903.46	77$\pm$6
17.06	5342.38	62$\pm$6
	adopted	63$\pm$6
	N2+
3	5666.64	61$\pm$6
	5676.02	48$\pm$7
	5679.56	49$\pm$5
	5686.21	53$\pm$8
	5710.76	51$\pm$8
	5730.65	69:
	adopted	52$\pm$4
	O2+
1	4638.85	89$\pm$11
	4641.81	80$\pm$7
	4649.14	81$\pm$6
	4650.84	87$\pm$11
	4661.64	83$\pm$9
	4676.24	61$\pm$9
	sum	81$\pm$7
2b	4317.14d	84$\pm$18
	4319.63	59$\pm$14
	4325.76c	242$\pm$73
	4336.83	76$\pm$28
	4345.56	103$\pm$19
	4349.43	65$\pm$9
	4366.89	116$\pm$19
	sume	81$\pm$17
10b	4072.15	81$\pm$9
	4092.93	66:
20	4110.79	121$\pm$8
	4019.22	59$\pm$12
	average	81$\pm$12
	adopted	81$\pm$12
	Ne2+
1	3694.22	21$\pm$6
	3709.62	27$\pm$9
	3766.26	25$\pm$10
	3777.14	27$\pm$11
	sum	24$\pm$6
39	3829.77	18:
55	4409.30	11$\pm$4
57	4428.77	4:
	adopted	19$\pm$6
	Mg2+
1	4481.21	3.6$\pm$0.8
a Blended with [Kr III] $\lambda$9902.30
b Assuming case A.
c Blended (see table LABEL:lineid_uves).
d Affected by charge transfer.
e Without considering blended lines.

Table 10. Table 12: Total abundances (12 + log(X/H)).

					PE04a			PO02b			T03, T04c
Element	CELs sum	CELs ICF	ORLs sum	ORLs ICF	CELs	ORLs		CELs	ORLs		CELs	ORLs	Model
He	–	–	11.11$\pm$0.03	–	–	11.09		–	11.09		–	11.08	11.03
C	–	–	8.83 $\pm$0.05h	8.89$\pm$0.10	–	8.85		8.64	–		8.33	8.86	8.42
N	–	8.52${}_{-0.21}{}^{}{}_{}{}^{+0.30}$d	–	8.79${}_{-0.08}{}^{}{}_{}{}^{+0.16}$i	–	8.82		8.66	–		8.52	8.77	8.39
O	8.78$\pm$0.05	–	8.98$\pm$0.06	–	8.63	8.87		8.72	–		8.79	9.10	8.80
Ne	–	8.40$\pm$0.12	–	–	8.05	–		8.20	–		8.30	–	8.09
Mg	–	–	7.56$\pm$0.10	–	–	–		–	–		–	–	–
P	–	5.06:	–	–	–	–		–	–		–	–	5.30
S	7.18$\pm$0.03	–	–	–	7.34	–		7.08	–		7.31	–	7.27
Cl	5.35$\pm$0.07	–	–	–	5.36	–		–	–		5.41	–	5.37
Ar	–	6.73$\pm$0.22	–	–	6.56	–		6.66	–		6.56	–	6.49
Fe	5.40$\pm$0.10	5.77 $\pm$0.22/5.48$\pm$0.17g	–	–	–	–		–	–		–	–	5.70
Se	–	3.60$\pm$0.17e	–	–	–	–		–	–		–	–	–
Kr	–	3.60$\pm$0.07f	–	–	–	–		–	–		–	–	–
Xe	–	3.43:	–	–	–	–		–	–		–	–	–
Br	–	3.53:	–	–	–	–		–	–		–	–	–
Rb	–	2.87	–	–	–	–		–	–		–	–	–
a Peimbert et al. (2004)
b Pottasch et al. (2002)
c Tsamis et al. (2003, 2004)
d Using the ICF N/O=N+/O+
e Using the ICFs given by eq. 9 of Sterling et al. (2015)
f Using the ICF given by eq. 4 of Sterling et al. (2015)
g Using the ICFs of eq. 2/eq. 3 provided by Rodríguez & Rubin (2005)
h Scaling C2+/C+ from CELs using an average value between Dufour et al. (2015) and Pottasch et al. (2002) data
i Assuming N/O = N2+/O2+ (ORLs)

Table 11. Table 13: Kr and Se total abundances (12 + log(X/H)).

ICFa	Kr	Se
eq. 1	3.61$\pm$0.07	–
eq. 2	3.84$\pm$0.24	–
eq. 3	4.17$\pm$0.09	–
eq. 4	3.60$\pm$0.07	–
eq. 7	–	3.89$\pm$0.28
eq. 8	–	3.13$\pm$0.11
eq. 9	–	3.60$\pm$0.17
a ICFs provided by Sterling et al. (2015)

Table 12. Table 14: N/O and C/O ratios in NGC 5315.

N/O	C/O	Reference
0.59${}_{-0.15}{}^{}{}_{}{}^{+0.24}$	0.76$\pm$0.13	This paper
0.41	0.64	Dufour et al. (2015)
0.89	–	Milingo et al. (2010)
0.93	0.95	Peimbert et al. (2004)
0.47	0.58	Tsamis et al. (2004)
0.54	0.35	Tsamis et al. (2003)
0.88	0.85	Pottasch et al. (2002)
0.48	–	Liu et al. (2001)
0.59	–	Samland et al. (1992)
0.36	–	de Freitas Pacheco et al. (1991)

Table 13. Table 15: Neutron-Capture Element Abundances.

	NGC 5315	NGC 3918a	Type I PNe$>$b	non-Type I PNe$>$b
$[$Se/(O,Ar)$]$	-0.07$\pm$0.29	0.19	-0.01	0.20
$[$Kr/(O,Ar)$]$	0.02$\pm$0.26	0.68	0.13	0.91
$[$Br/(O,Ar)$]$	0.66:	–	–	–
$[$Rb/(O,Ar)$]$	0.02	$>$0.00	–	–
$[$Xe/(O,Ar)$]$	0.86:	$>$0.26	–	–
a García-Rojas et al. (2015)
b Mean values from the sample of Sterling et al. (2015)

Table 14. Table 2: Observed and reddening corrected line ratios (F(H β 𝛽 \beta ) = 100) and line identifications in NGC 5315.

${\lambda }_0$ (Å )	Ion	Mult.	${\lambda }_{\mathrm{obs}}$	$V_{\mathrm{rad}}$ (km s-1)	$F\left(\lambda \right)/F\left(Href\right)$d	$I\left(\lambda \right)/I\left(H\beta \right)$e	Err ($\%$)f	notes
3109.18	[Ar iii]	2F	3108.70	-46.28	0.132	0.276	14
3187.84	He i	3	3187.31	-49.84	1.240	2.453	10
3218.19	Ne ii	4D0-4F	3217.71	-44.71	0.0741	0.144	16
3244.00	O ii	4P-D[2]0	3243.40	-55.45	0.0615	0.117	18	?
3258.27	He i	1S-1P0	3257.79	-44.16	0.0401	0.0757	22
3296.77	He i	9	3296.22	-50.01	0.0363	0.0671	23
3323.74	Ne ii	7	3323.32	-37.88	0.0287	0.0524	27
3328.72	N ii	3D-3P0	3328.39	-29.72	0.0461	0.0839	20
3331.31	N ii	3D-3P0	3330.81	-45.00	0.0175	0.0318	36
3334.87	Ne ii	2	3334.35	-46.75	0.0885	0.161	15
3342.50	[Ne iii]	2F	3342.30	-17.94	0.103	0.186	14
3342.85	[Cl iii]	2F	*	*	*	*	*
3353.21	[Cl iii]	2F	3352.73	-42.91	0.0293	0.0527	23
3354.42	He i	8	3354.20	-19.66	0.127	0.228	14
3367.05	Ne ii	12	3366.73	-28.49	0.0255	0.0455	28
	?		3417.17		0.0191	0.0334	34
3418.83	[Mn iii]	a6S-a4P	*		*	*	*
3422.82	[Mn iii]	a6S-a4P	3422.37	-39.41	0.0162	0.0283	37
3427.47	[Mn iii]	a6S-a4P	3427.07	-34.99	0.0267	0.0466	27
3434.29	He i	3P0-3D	3433.82	-41.03	0.0086	0.0150	:
3435.21	He i	3P0-3D	3434.72	-42.76	0.0098	0.0170	:
3436.22	He i	3P0-3D	3435.73	-42.75	0.0135	0.0236	:
3437.34	He i	3P0-3D	3436.83	-44.48	0.0197	0.0342	33
3438.71	He i	3P0-3D	3438.26	-39.23	0.0133	0.0231	:
3439.99	He i	3P0-3D	3439.56	-37.47	0.0115	0.0200	:
3441.56	He i	3P0-3D	3441.12	-38.33	0.0110	0.0192	:
3443.33	He i	3P0-3D	3442.88	-39.18	0.0138	0.0240	:
3445.33	He i	3P0-3D	3444.92	-35.68	0.0167	0.0290	36
3447.63	He i	3P0-3D	3447.10	-46.09	0.185	0.321	12
3450.26	He i	3P0-3D	3449.75	-44.31	0.0193	0.0334	33
3453.30	He i	3P0-3D	3452.85	-39.07	0.0242	0.0418	29
3456.84	He i	3P0-3D	3456.31	-45.96	0.0287	0.0497	27
3460.99	He i	3P0-3D	3460.51	-41.58	0.0277	0.0478	27
3465.92	He i	3P0-3D	3465.43	-42.38	0.0383	0.0660	23
3466.50	[N i]	4S0-2P0	3466.04	-39.78	0.0429	0.0738	21
3471.80	He i	44	3471.34	-39.72	0.0487	0.0836	20
3478.97	He i	43	3478.49	-41.36	0.0494	0.0847	20
3487.73	He i	42	3487.24	-42.12	0.0655	0.112	17
3498.64	He i	40	3498.15	-41.99	0.0683	0.116	17
3512.51	He i	38	3512.02	-41.82	0.0963	0.163	15
3530.50	He i	36	3530.00	-42.46	0.147	0.248	13
3554.42	He i	34	3553.92	-42.17	0.198	0.332	12
3562.95	He i	33	3562.49	-38.71	0.0118	0.0198	:
3568.50	Ne ii	5	3568.02	-40.33	0.0262	0.0438	28
3574.61	Ne ii	5	3574.07	-45.29	0.0107	0.0178	:
3587.28	He i	31	3586.78	-41.79	0.267	0.444	12
3590.86	C ii	23	3590.07	-65.96	0.0178	0.0294	35
3599.32	He i	30	3598.82	-41.65	0.0182	0.0302	34
3613.64	He i	6	3613.12	-43.14	0.318	0.524	11
3634.25	He i	28	3633.73	-42.90	0.381	0.624	11
3669.47	H i	H25	3668.94	-43.30	0.305	0.495	12
3671.48	H i	H24	3670.95	-43.28	0.350	0.568	11
3673.76	H i	H23	3673.24	-42.43	0.401	0.650	11
3676.37	H i	H22	3675.84	-43.22	0.456	0.739	11
3679.36	H i	H21	3678.83	-43.18	0.517	0.836	11
3682.81	H i	H20	3682.28	-43.14	0.575	0.929	11
3686.83	H i	H19	3686.30	-43.10	0.666	1.075	11
3691.56	H i	H18	3691.03	-43.04	0.739	1.191	11
3694.22	Ne ii	1	3693.69	-43.01	0.0474	0.0764	20
3697.15	H i	H17	3696.62	-42.98	0.847	1.364	10
3703.86	H i	H16	3703.34	-42.09	0.910	1.462	10
3705.04	He i	25	3704.48	-45.31	0.582	0.935	11
3709.62	Ne ii	1	3709.09	-42.83	0.0227	0.0364	30
3711.97	H i	H15	3711.48	-39.57	1.172	1.879	10
3713.08	Ne ii	5	*	*	*	*	*
3717.72	S iii	6	3717.26	-37.09	0.0144	0.0231	:
3721.83	[S iii]	2F	3721.27	-45.11	2.571	4.110	10
3721.93	H i	H14	*	*	*	*	*
3726.03	[O ii]	1F	3725.33	-56.32	20.54	32.80	10
3728.82	[O ii]	1F	3728.01	-65.12	8.244	13.15	10
3734.37	H i	H13	3733.81	-44.96	1.520	2.421	10
3737.55	Ne ii	4D-2[2]0	3737.15	-32.08	0.0113	0.0180	:
3745.94	He i	1P0-1D	3745.39	-44.02	0.0132	0.0210	:
3750.15	H i	H12	3749.60	-43.97	2.029	3.216	10
3756.10	He i	66	3755.57	-42.30	0.0158	0.0250	38
3762.47	O ii	4S0-4P	3761.94	-42.23	0.0075	0.0119	:
3766.26	Ne ii	1	3765.78	-38.21	0.0171	0.0270	36
3768.78	He i	65	3770.08	103.41	2.498	3.935	10
3770.63	H i	H11	*	*	*	*	*
3777.14	Ne ii	1	3776.67	-37.30	0.0178	0.0280	35
3784.89	He i	64	3784.31	-45.94	0.0275	0.0431	27
3797.90	H i	H10	3797.35	-43.42	3.454	5.394	10
3797.63	[S iii]	2F	*	*	*	*	*
3805.74	He i	63	3805.39	-27.57	0.0620	0.0966	18
3819.61	He i	22	3819.07	-42.38	1.061	1.646	10
3829.77	Ne ii	39	3829.21	-43.84	0.0134	0.0207	:
3831.72	C ii	96	3831.08	-50.07	0.0310	0.0479	25
3835.39	H i	H9	3834.83	-43.77	4.891	7.547	10
3853.66	Si ii	1	3853.17	-38.12	0.0070	0.0107	61
3856.02	Si ii	1	3855.47	-42.76	0.0769	0.118	16
3862.59	Si ii	1	3862.07	-40.36	0.0479	0.0732	20
3868.75	[Ne iii]	1F	3868.15	-46.49	43.69	66.67	9
3880.33	Ar ii	54	3879.62	-54.85	0.0310	0.0471	22
3882.19	O ii	12	3881.73	-35.52	0.0172	0.0262	31
3888.65	He i	2	3888.34	-23.90	11.99	18.17	7
3889.05	H i	H8	*	*	*	*	*
	?		3900.55		0.0483	0.0728	17
3918.98	C ii	4	3918.52	-35.19	0.0253	0.0379	25
3920.68	C ii	4	3920.21	-35.94	0.0437	0.0655	18
3926.53	He i	58	3925.96	-43.52	0.114	0.170	11
3935.94	He i	57	3935.40	-41.13	0.0061	0.0091	:
3964.73	He i	5	3964.12	-46.13	0.660	0.973	8
3967.46	[Ne iii]	1F	3966.89	-43.07	13.35	19.66	7
3970.07	H i	H7	3969.47	-45.31	10.26	15.09	7
3973.24	O ii	6	3972.71	-39.99	0.0150	0.0220	34
3994.98	N ii	12	3994.41	-42.77	0.0327	0.0477	21
4009.26	He i	55	4008.64	-46.36	0.182	0.264	10
4023.98	He i	54	4023.40	-43.21	0.0175	0.0252	31
4026.21	He i	18	4025.60	-45.42	2.108	3.034	7
4035.07	O ii	68	4034.53	-40.12	0.0140	0.0200	36
4035.08	N ii	39a	*	*	*	*	*
4041.31	N ii	39b	4040.71	-44.51	0.0376	0.0538	20
4043.53	N ii	39a	4042.91	-45.97	0.0122	0.0175	39
4060.60	O ii	97	4060.19	-30.27	0.0147	0.0209	35
4062.94	O ii	50	4062.27	-49.44	0.0065	0.0091	:
4068.60	[S ii]	1F	4067.99	-44.95	4.374	6.189	7
4069.62	O ii	10	*	*	*	*	*
4069.89	O ii	10	*	*	*	*	*
4072.15	O ii	10	4071.57	-42.70	0.118	0.167	11
4075.86	O ii	10	4075.68	-49.27	1.600	2.257	7
4076.35	[S ii]	1F	*	*	*	*	*
4087.15	O ii	48	4086.53	-45.48	0.0118	0.0166	:
4089.29	O ii	48	4088.64	-47.65	0.0577	0.0809	16	a
4092.93	O ii	10	4092.33	-43.95	0.0132	0.0185	38
4097.26	O ii	48	4096.69	-41.71	0.204	0.285	9
4101.74	H i	H6	4101.12	-45.32	17.43	24.33	7
4110.79	O ii	20	4110.10	-50.32	0.0211	0.0293	27
4119.22	O ii	20	4118.59	-45.85	0.0376	0.0521	20
4120.82	He i	16	4120.20	-45.11	0.287	0.397	8	a
4129.32	O ii	19	4128.73	-42.83	0.0069	0.0095	:
4132.80	O ii	19	4132.12	-49.33	0.0426	0.0587	18
4143.76	He i	53	4143.13	-45.58	0.308	0.422	9
4145.90	O ii	106	4145.43	-33.99	0.0207	0.0283	28
4146.08	O ii	106	*	*	*	*	*
4153.30	O ii	19	4152.65	-46.92	0.0524	0.0715	16
4156.53	O ii	19	4155.96	-41.11	0.0720	0.0982	14	a
4168.97	He i	52	4168.42	-39.55	0.0471	0.0639	17
4171.61	N ii	43a	4170.99	-44.56	0.0035	0.0047	:
4176.16	N ii	43a	4175.56	-43.07	0.0091	0.0123	:
4185.45	O ii	36	4184.82	-45.13	0.0313	0.0422	22
4189.79	O ii	36	4189.17	-44.36	0.0456	0.0613	18
4219.76	Ne ii	52a	4219.12	-45.47	0.0117	0.0156	:
4231.53	Ne ii	52b	4230.94	-41.80	0.0098	0.0129	:
4233.27	Ni ii	2G0-2F	4232.51	-53.82	0.0250	0.0329	25
4236.91	N ii	48	4236.41	-35.38	0.0313	0.0412	22
4237.05	N ii	48	*	*	*	*	*
4241.78	N ii	48	4241.15	-44.53	0.0313	0.0410	22
4243.97	[Fe ii]	21F	4243.40	-40.26	0.0112	0.0147	:
4254.00	O ii	101	4253.21	-55.67	0.0332	0.0434	21
4267.15	C ii	6	4266.43	-50.58	0.551	0.715	8
4275.55	O ii	67	4275.69	9.82	0.0905	0.117	12
4276.75	O ii	67	*	*	*	*	*	a
4281.32	O ii	53b	4280.75	-39.91	0.0058	0.0075	:
4282.96	O ii	67c	4282.57	-27.30	0.0118	0.0152	:
4285.69	O ii	78	4284.91	-54.56	0.0098	0.0126	:
4287.39	[Fe ii]	7F	4286.80	-41.26	0.0128	0.0164	38
4291.25	O ii	55	4290.58	-46.81	0.0058	0.0074	:
4292.21	O ii	78	4291.58	-44.00	0.0063	0.0080	:
4294.92	O ii	54	4294.15	-53.75	0.0157	0.0201	33
4303.61	O ii	65	4303.14	-32.74	0.0226	0.0288	27
4303.82	O ii	53	*	*	*	*	*
4307.23	O ii	54	4306.61	-43.15	0.0044	0.0056	:
4309.00	O ii	4D-D[1]0	4308.39	-42.44	0.0041	0.0052	:
4309.01	O ii	4D-D[1]0	*	*	*	*	*
4317.14	O ii	2	4316.51	-43.75	0.0338	0.0428	21	a
4319.63	O ii	2	4319.00	-43.72	0.0241	0.0306	25
4325.76	O ii	2	4325.19	-39.50	0.0192	0.0242	30
	?		*		*	*	*
4329.75	C ii	2D-2F0	4329.14	-41.54	0.0027	0.0033	:
4332.71	O ii	65	4332.13	-40.13	0.0063	0.0079	:
4336.83	O ii	2	4336.12	-49.08	0.0128	0.0161	37
4340.47	H i	H5	4339.81	-45.59	37.71	47.33	7
4345.56	O ii	2	4344.94	-42.08	0.0456	0.0571	18
4346.85	[Fe ii]	21F	4346.77	-5.52	0.0059	0.0074	:
4349.43	O ii	2	4348.77	-45.49	0.0732	0.0914	14
4359.34	[Fe ii]	7F	4358.61	-50.20	0.0120	0.0150	39
4363.21	[O iii]	2F	4362.58	-43.29	3.045	3.783	7
4366.89	O ii	2	4366.21	-46.68	0.0532	0.0659	16
4368.19	O i	5	4367.64	-37.75	0.0581	0.0720	15
4368.25	O i	5	*	*	*	*	*
4371.59	C ii	45	4371.26	-22.63	0.0118	0.0146	40
4379.55	Ne ii	60	4378.97	-39.70	0.0072	0.0089	:
4387.93	He i	51	4387.26	-45.78	0.630	0.773	8
4391.94	Ne ii	55e	4391.61	-22.53	0.0400	0.0490	19	a
4409.30	Ne ii	55e	4408.67	-42.83	0.0194	0.0236	29
4414.90	O ii	5	4414.27	-42.78	0.0331	0.0402	22
4416.76	Ne ii	2D-2[2]0	4416.21	-37.33	0.0437	0.0530	18
4416.97	O ii	5	*	*	*	*	*
4428.54	Ne ii	57	4427.75	-53.48	0.0179	0.0215	31
4430.94	Ne ii	61	4430.28	-44.65	0.0056	0.0067	:
4432.74	N ii	55a	4432.14	-40.58	0.0331	0.0398	21
4437.55	He i	50	4436.89	-44.59	0.0626	0.0752	15
4439.46	Ar ii	2D0-2D	4438.78	-45.92	0.0061	0.0074	:
4439.88	Ar ii	2D0-2D	*	*	*	*	*
4442.69	Ne ii	60	4441.83	-58.03	0.0120	0.0144	39
4447.03	N ii	1P-1D0	4446.28	-50.56	0.0053	0.0063	:
4448.46	Ar ii	2D0-2D	4447.86	-40.44	0.0088	0.0105	:
4448.88	Ar ii	2D0-2D	*	*	*	*	:
4452.38	O ii	5	4451.59	-53.19	0.0127	0.0152	38
4457.05	Ne ii	2D-2[2]0	4456.75	-20.18	0.0121	0.0144	39
4457.26	Ne ii	2D-2[2]0	*	*	*	*	*
	?		4461.27		0.0038	0.0045	:	?
4463.65	O i	3P-3P0	4462.84	-54.40	0.0051	0.0061	:
4463.66	O i	3P-3P0	*	*	*	*	*
4465.41	O ii	94	4464.76	-43.64	0.0196	0.0232	29
4467.92	O ii	94	4467.21	-47.64	0.0088	0.0104	:
4471.47	He i	14	4470.83	-42.91	5.317	6.286	7
4481.21	Mg ii	4	4480.74	-31.44	0.0316	0.0372	22
4483.43	S ii	4D0-4P	4482.71	-48.14	0.0032	0.0038	:
4491.23	O ii	86a	4490.57	-44.06	0.0161	0.0189	33
4498.92	Ne ii	4P-2[1]0	4498.41	-33.98	0.0083	0.0096	:
4499.12	Ne ii	4P-2[1]0	*	*	*	*	*
4520.69	Si ii	2D-2P0	4520.05	-42.44	0.0182	0.0211	30
4525.76	Ne i	–	4524.90	-56.97	0.0081	0.0094	:
4530.41	N ii	58b	4529.43	-64.85	0.0524	0.0603	16
4530.86	N iii	3	*	*	*	*	*
4552.52	N ii	58a	4552.05	-30.95	0.0203	0.0231	29
4562.60	Mg i]	1	4561.72	-57.82	0.0134	0.0152	37
4571.10	Mg i]	1	4570.36	-48.53	0.217	0.246	9	a
4590.97	O ii	15	4590.30	-43.75	0.0532	0.0596	16
4595.95	O ii	15	4595.47	-31.31	0.0449	0.0502	18
4596.18	O ii	15	*	*	*	*	*
4601.48	N ii	5	4600.92	-36.48	0.0543	0.0606	16
4607.13	[Fe iii]	3F	4606.42	-46.20	0.0367	0.0408	20
4609.44	O ii	92a	4609.17	-17.56	0.0547	0.0608	16
4610.20	O ii	2D-F[2]0	*	*	*	*	*
4621.39	N ii	5	4620.68	-46.06	0.0242	0.0268	26
4630.54	N ii	5	4629.82	-46.61	0.132	0.145	11
4634.14	N iii	2	4633.46	-43.99	0.110	0.121	11
4638.86	O ii	1	4638.17	-44.59	0.0920	0.101	12
4640.64	N iii	2	4639.94	-45.22	0.210	0.230	9
4641.81	O ii	1	4641.11	-45.21	0.202	0.221	9
4643.06	N ii	5	4642.29	-49.72	0.0622	0.0681	15
4649.13	O ii	1	4648.44	-44.49	0.367	0.400	8
4650.84	O ii	1	4650.13	-45.77	0.0845	0.0922	13
4658.05	[Fe iii]	3F	4657.32	-46.98	0.227	0.247	9
4661.63	O ii	1	4660.92	-45.66	0.0973	0.106	12	a
4673.73	O ii	1	4672.97	-48.75	0.0124	0.0134	39
4676.24	O ii	1	4675.54	-44.88	0.0554	0.0598	17
4699.22	O ii	25	4698.51	-45.30	0.0214	0.0229	28
4701.62	[Fe iii]	3F	4700.80	-52.29	0.0743	0.0793	14
4705.35	O ii	25	4704.67	-43.32	0.0119	0.0127	:
4711.37	[Ar iv]	1F	4710.94	-27.36	0.0634	0.0674	12
4713.14	He i	12	4712.47	-42.62	0.698	0.741	8
4733.93	[Fe iii]	3F	4733.13	-50.66	0.0152	0.0160	34
4740.17	[Ar iv]	1F	4739.56	-38.58	0.218	0.229	9
4754.69	[Fe iii]	3F	4753.98	-44.77	0.0176	0.0183	31
4769.43	[Fe iii]	3F	4768.67	-47.77	0.0184	0.0191	30
4777.68	[Fe iii]	3F	4776.93	-47.06	0.0101	0.0105	:
4788.13	N ii	20	4787.42	-44.45	0.0148	0.0152	34
4802.70	C ii	17.08	4802.51	-11.86	0.0221	0.0226	27
4814.55	[Fe ii]	20F	4813.89	-41.10	0.0080	0.0081	:
4815.51	S ii	9	4814.85	-41.09	0.0071	0.0072	:
4861.33	H i	H4	4860.59	-45.63	100.0	100.0	7
4881.00	[Fe iii]	2F	4880.21	-48.52	0.128	0.127	11
4921.93	He i	48	4921.19	-45.07	1.701	1.662	7
4931.32	[O iii]	1F	4930.49	-50.46	0.100	0.0977	12
4958.91	[O iii]	1F	4958.17	-44.74	292.2	281.6	6
4980.05	O i	3P-3S0	4979.30	-47.56	0.0108	0.0103	19
4987.20	[Fe iii]	2F	4986.62	-34.87	0.0335	0.0319	11
4987.38	N ii	24	*	*	*	*	*
4994.37	N ii	24	4993.68	-41.42	0.0757	0.0720	8
5006.84	[O iii]	1F	5006.09	-44.91	892.8	845.5	6
5015.68	He i	4	5014.89	-47.22	3.195	3.016	6
5025.66	N ii	19	5025.00	-39.37	0.0102	0.0096	20
5032.13	C ii	2P-2D	5031.29	-50.04	0.0614	0.0576	9
5035.94	C ii	2P-2D	5035.07	-51.79	0.0380	0.0356	10
5041.03	Si ii	5	5040.29	-44.01	0.178	0.167	8
5045.10	N ii	4	5044.44	-39.22	0.0678	0.0633	9
5048.18	[Fe ii]	a2P-c2D	5047.34	-49.88	0.641	0.598	7
5055.98	Si ii	5	5055.39	-34.98	0.110	0.103	8
5121.82	C ii	12	5121.09	-42.73	0.0177	0.0161	14
5145.16	C ii	4P0-4P	5144.38	-45.45	0.0033	0.0030	:
5146.61	O i	28	5145.97	-37.28	0.0099	0.0089	20
5146.65	O i	28	*	*	*	*	*
5158.00	[Fe ii]	18F	5157.11	-51.73	0.0097	0.0087	20
5158.81	[Fe ii]	19F	5158.05	-44.17	0.0318	0.0286	11
	?		5172.58		0.0056	0.005	28
5175.89	N ii	5D0-5F	5175.23	-38.23	0.0106	0.0095	19
5179.52	N ii	5D0-5F	5178.84	-38.20	0.0274	0.0245	12
5191.82	[Ar iii]	3F	5191.00	-47.35	0.147	0.131	8
5197.90	[N i]	1F	5197.10	-46.14	0.555	0.493	7
5200.06	[Fe ii]	19F	5199.43	-36.32	0.365	0.324	7
5200.26	[N i]	1F	*	*	*	*	*
5261.61	[Fe ii]	19F	5260.92	-39.31	0.0173	0.0151	15
5270.40	[Fe iii]	1F	5269.71	-39.25	0.104	0.0906	8
5274.97	O i	27	5274.40	-32.39	0.0064	0.0055	26
5275.12	O i	27	*	*	*	*	*
5298.89	O i	26	5298.33	-31.68	0.0256	0.0220	12
5299.04	O i	26	*	*	*	*	*
5333.65	[Fe ii]	19F	5332.95	-39.35	0.0090	0.0077	21
5342.38	C ii	17.06	5341.61	-43.21	0.0414	0.0352	10
5346.02	[Kr iv]	4S-2D	5345.27	-42.06	0.0096	0.0081	20
	?		5358.44		0.0103	0.0087	19
	?		*		*	*	*
5367.73	C ii	4P0-4D	5367.22	-28.48	0.0062	0.0052	27
5370.39	C ii	4P0-4D	5369.57	-45.78	0.0055	0.0046	29
5376.45	[Fe ii]	19F	5375.81	-35.69	0.0170	0.0143	15
5405.15	Ne ii	2[5]0-2[6]	5404.49	-36.61	0.0061	0.0051	27	?
5412.00	[Fe iii]	1F	5411.13	-48.19	0.0207	0.0173	13
5433.13	[Fe ii]	18F	5432.16	-53.52	0.0115	0.0096	19
5452.07	N ii	29	5451.32	-41.24	0.0121	0.0100	18
5454.22	N ii	29	5453.33	-48.92	0.0289	0.0238	12
5462.58	N ii	29	5461.90	-37.32	0.0231	0.0189	13
5478.09	N ii	29	5477.30	-43.23	0.0069	0.0056	25
5480.05	N ii	29	5479.22	-45.41	0.0142	0.0116	16
5495.98	N ii	29	5494.90	-58.91	0.0382	0.0311	10
5506.87	[Cr iii]	2F	5506.09	-42.46	0.0034	0.0027	:	?
5512.77	O i	25	5511.98	-42.96	0.0113	0.0092	19
5517.71	[Cl iii]	1F	5516.91	-43.47	0.395	0.320	7
5526.23	N ii	63	5525.49	-40.14	0.0094	0.0076	21
5527.34	[Fe ii]	17F	5526.56	-42.31	0.0145	0.0118	16
5530.24	N ii	63	5529.49	-40.66	0.0165	0.0133	15
5535.36	N ii	63	5534.59	-41.70	0.0335	0.0270	11
5537.88	[Cl iii]	1F	5537.06	-44.39	1.151	0.928	7
5543.47	N ii	63	5542.71	-41.10	0.0112	0.0090	19
5551.96	N ii	63	5551.20	-41.04	0.0130	0.0104	17	?
5577.34	[O i]	3F	5576.61	-39.24	0.0589	0.0470	9
5640.19	S ii	4F-4D0	5639.30	-47.31	0.0087	0.0068	22
5666.64	N ii	3	5665.84	-42.32	0.1371	0.107	8
5676.02	N ii	3	5675.22	-42.25	0.0495	0.0386	10
5679.56	N ii	3	5678.78	-41.17	0.244	0.190	7
5686.21	N ii	3	5685.43	-41.12	0.0321	0.0250	11
5710.76	N ii	3	5709.97	-41.47	0.0414	0.0320	10
5730.65	N ii	3	5729.79	-44.99	0.0032	0.0025	:
5739.73	Si iii	4	5739.02	-37.08	0.0074	0.0057	24
5747.33	O ii	2D0-2F	5746.52	-42.25	0.0107	0.0082	19
5754.64	[N ii]	3F	5753.79	-44.28	6.074	4.648	7
5867.74	[Kr iv]	4S-2D	5867.02	-36.79	0.0252	0.0188	12
5875.60	He i	11	5874.83	-39.29	26.75	19.93	7
6151.43	C ii	16.04	6150.41	-49.71	0.0385	0.0271	10
6300.30	[O i]	1F	6299.38	-43.78	7.740	5.289	7	b
6312.10	[S iii]	3F	6311.20	-42.75	4.930	3.361	7
6340.58	N ii	3D0-3P	6339.64	-44.44	0.0044	0.0030	34
6347.11	Si ii	2	6346.21	-42.51	0.151	0.102	8
6363.78	[O i]	1F	6362.85	-43.81	2.692	1.817	7	b
6371.36	Si ii	2	6370.49	-40.94	0.110	0.0741	8
6402.25	Ne i	1	6401.33	-43.08	0.0160	0.0107	15
6461.95	C ii	17.04	6460.86	-50.57	0.0993	0.0658	8
6482.05	N ii	1P0-1P	6481.18	-40.24	0.0291	0.0192	11
6486.46	O ii	G[3]0-1[4]	6485.51	-43.91	0.0029	0.0019	:
6527.24	[N ii]	1F	6526.25	-39.04	0.0468	0.0306	10
6548.03	[N ii]	1F	6546.95	-49.45	87.42	56.95	6
6555.56	[Br iii]	–	6554.99	-26.07	0.0149	0.0097	:	?
6562.82	H i	H3	6561.74	-49.33	436.9	283.8	6
6578.05	C ii	2	6576.93	-51.04	0.464	0.301	6
6583.41	[N ii]	1F	6582.32	-49.64	275.7	178.4	6
6610.56	N ii	31F	6609.65	-41.27	0.0088	0.0056	22
6641.03	O ii	43a	6640.36	-30.25	0.0078	0.0050	24
6678.15	He i	46	6677.21	-42.20	8.195	5.206	7
6716.47	[S ii]	2F	6715.37	-49.10	7.088	4.470	10
6730.85	[S ii]	2F	6729.76	-48.55	14.32	9.004	10
6779.93	C ii	14	6780.47	23.88	1.779	0.111	11
6780.59	C ii	14	*	*	*	*	*
6791.47	C ii	14	6790.49	-43.26	0.104	0.0065	30
6803.01	[Cr ii]	b4D-c2F	6802.24	-33.93	0.805	0.0503	12	?
6809.99	N ii	54	6808.98	-44.46	0.0853	0.0053	34
6826.70	[Kr iii]	3P-1D	6826.27	-18.88	0.344	0.0215	16
6827.88	He i	1/16	*	*	*	*	*
6855.88	He i	1/15	6856.06	7.87	0.511	0.0319	14	a
6933.89	He i	1/13	6932.98	-39.34	0.151	0.0094	24
6989.47	He i	1/12	6988.46	-43.32	0.161	0.0100	23
7001.92	O i	21	7001.14	-33.40	0.883	0.0547	12
7062.26	He i	1/11	7064.24	84.05	136.2	8.411	10
7065.28	He i	10	*	*	*	*	*
7135.78	[Ar iii]	1F	7134.76	-42.85	500.6	30.81	10
	?		7152.71		0.161	0.0099	16
7155.16	[Fe ii]	14F	7154.16	-41.90	1.946	0.120	11
7160.61	He i	1/10	7159.52	-45.63	0.340	0.0210	16
7231.34	C ii	3	7230.08	-52.24	2.514	0.154	11
7236.42	C ii	3	7235.36	-43.91	3.823	0.234	11	c
7237.17	C ii	3	*		*	*	*
7254.15	O i	3P-3S0	7253.36	-32.81	1.576	0.0965	11	b
7254.45	O i	3P-3S0	*		*	*	*
7254.53	O i	3P-3S0	*		*	*	*
7281.35	He i	45	7280.29	-43.64	13.81	0.845	10
7298.05	He i	1/9	7296.91	-46.83	0.610	0.0373	13
7318.92	[O ii]	2F	7318.81	-4.51	119.3	7.284	10	b
7319.99	[O ii]	2F	*		*	*	*
7329.66	[O ii]	2F	7329.12	-22.09	107.7	6.572	10	b
7330.73	[O ii]	2F	*		*	*	*
7377.83	[Ni ii]	2F	7376.86	-39.42	0.991	0.0603	12
7388.16	[Fe ii]	14F	7387.04	-45.45	0.218	0.0132	20
7423.64	N i	3	7422.56	-43.61	0.139	0.0085	25
7442.30	N i	3	7441.28	-41.09	0.262	0.0159	18
7452.54	[Fe ii]	14F	7451.17	-55.11	0.439	0.0267	15
	?		*		*	*	*
7468.31	N i	3	7467.26	-42.15	0.501	0.0304	14
7499.85	He i	1/8	7498.73	-44.77	0.822	0.0498	12
7504.96	O ii	2G-G[5]0	7503.85	-44.34	0.121	0.0073	28
7509.24	N ii	3P-1P	7508.46	-31.14	0.0511	0.0031	:
7519.49	C ii	16.08	7518.55	-37.48	0.152	0.0092	24
7519.86	C ii	16.08	*	*	*	*	*
7530.54	[Cl iv]	1F	7529.42	-44.59	0.627	0.0379	13
7535.40	[Xe iv]	4S-2D	7534.05	-53.71	0.0526	0.0032	:	?
7561.27	[Fe ii]	b4F-b2D	7560.30	-38.46	0.0454	0.0027	:	?
7751.10	[Ar iii]	2F	7750.00	-42.55	126.5	7.577	10
7771.93	O i	1	7770.83	-42.43	0.283	0.0169	18	b
7816.13	He i	1/7	7814.98	-44.11	1.335	0.0797	11
7875.99	[P ii]	1D-1S	7874.79	-45.68	0.275	0.0164	18	?
8045.63	[Cl iv]	1F	8044.64	-36.89	1.261	0.0746	12
8057.59	He i	4/18	8056.32	-47.25	0.168	0.0099	23
8084.29	He i	4/17	8083.05	-45.98	0.134	0.0079	26
8094.08	He i	2/10	8093.10	-36.30	0.158	0.0093	24
8116.30	He i	4/16	8115.30	-36.94	0.173	0.0102	22
8125.30	[Cr ii]	1F	8124.29	-37.27	0.0782	0.0046	36	?
8203.85	He i	4/14	8202.66	-43.49	0.266	0.0156	18
8216.34	N i	2	8215.04	-47.43	0.695	0.0408	13
8223.14	N i	2	8222.01	-41.20	0.579	0.0340	14	c
8267.94	H i	P34	8266.69	-45.32	1.147	0.0672	12
8271.93	H i	P33	8270.67	-45.67	1.352	0.0792	12
8276.31	H i	P32	8275.26	-38.03	0.991	0.0581	12	c
8286.43	H i	P30	8285.23	-43.41	1.747	0.102	11
8292.31	H i	P29	8291.12	-43.02	1.867	0.109	11
8298.83	H i	P28	8297.57	-45.52	2.324	0.136	11
8306.11	H i	P27	8304.84	-45.84	1.920	0.112	11	c
8314.26	H i	P26	8313.04	-43.99	2.514	0.147	11	c
8323.42	H i	P25	8322.34	-38.90	2.281	0.133	11	c
8329.87	He i	6/23	8328.74	-40.67	0.161	0.0094	23
8333.78	H i	P24	8332.55	-44.25	3.485	0.204	11
8343.27	He i	6/22	8344.30	37.01	4.774	0.279	11	b
8345.55	H i	P23	*	*	*	*	*
8359.00	H i	P22	8357.80	-43.04	4.140	0.242	11	c
8361.71	He i	68	8360.49	-43.74	2.218	0.130	11
8374.48	H i	P21	8373.28	-42.96	5.112	0.299	10
8392.40	H i	P20	8391.15	-44.65	5.218	0.304	11
8397.41	He i	6/19	8396.16	-44.63	0.128	0.0075	27
8413.32	H i	P19	8412.08	-44.19	6.063	0.354	11
8421.96	He i	6/18	8420.81	-40.94	0.222	0.0129	20
8433.85	[Cl iii]	3F	8432.47	-49.05	0.438	0.0255	15
8437.96	H i	P18	8436.72	-44.06	6.253	0.364	11
8446.25	O i	4	8445.31	-33.36	15.48	0.902	10
8446.36	O i	4	*	*	*	*	*
8446.76	O i	4	*	*	*	*	*
8451.16	He i	6/17	8449.96	-36.89	0.171	0.0010	23
8453.61	He i	7/17	*	*	*	*	*
8480.90	[Cl iii]	3F	8479.63	-44.89	0.547	0.0318	14
8486.27	He i	6/16	8485.01	-44.51	0.357	0.0208	16
8499.70	[Cl iii]	3F	8501.15	51.14	9.632	0.560	10
8502.48	H i	P16	*	*	*	*	*
8528.99	He i	6/15	8527.78	-42.53	0.429	0.0249	15
8531.51	He i	7/15	8530.20	-46.17	0.0699	0.0041	38
8648.27	He i	6/13	8647.04	-42.64	0.525	0.0304	14	b
8650.83	He i	7/13	8649.76	-37.08	0.262	0.0151	18	b
8665.02	H i	P13	8663.77	-43.25	13.81	0.799	10	b
8680.28	N i	1	8679.06	-42.14	0.408	0.0236	15
8683.40	N i	1	8682.19	-41.77	0.463	0.0267	15	b
8686.15	N i	1	8684.96	-41.07	0.389	0.0225	16
8703.25	N i	1	8701.99	-43.40	0.224	0.0129	20
8711.70	N i	1	8710.48	-41.98	0.325	0.0188	17
8718.84	N i	1	8717.61	-41.95	0.149	0.0086	25
8727.13	[C i]	3F	8725.83	-44.66	0.934	0.0539	12
8728.90	N i	1	*	*	*	*	*
8733.43	He i	6/12	8732.16	-43.60	0.731	0.0422	13
8736.04	He i	7/12	8734.69	-46.33	0.230	0.0133	19
8740.10	He i	5/12	8738.91	-40.82	0.0475	0.0027	:
8750.47	H i	P12	8749.19	-43.85	19.18	1.106	10
8776.60	He i	4/9	8775.60	-34.16	1.134	0.0654	12	b
8816.50	He i	10/12	8815.35	-39.10	0.0957	0.0055	31
8829.40	[S iii]	1F	8827.67	-58.74	1.265	0.0728	12	b
8845.38	He i	6/11	8844.06	-44.74	1.083	0.0623	12
8848.05	He i	7/11	8846.77	-43.37	0.589	0.0339	14	b
8854.20	He i	5/11	8852.74	-49.43	0.180	0.0103	25
8855.28	[Se iii]	3P-1D	*	*	*	*	*
8862.79	H i	P11	8861.49	-43.97	24.29	1.396	10
8873.40	C ii	2D-2F0	8872.10	-43.92	0.0454	0.0026	:	?
8891.91	[Fe ii]	13F	8890.58	-44.84	0.663	0.0381	12	b
8996.99	He i	6/10	8995.66	-44.32	1.472	0.0843	12
8999.40	He i	7/10	8998.34	-35.31	0.501	0.0287	14
9014.91	H i	P10	9013.63	-42.57	29.78	1.705	11	c
9063.29	He i	4/8	9061.94	-44.65	1.380	0.0789	9
9068.60	[S iii]	1F	9067.45	-38.02	1145.4	65.49	8	c
9085.42	He i	10/10	9084.05	-45.21	0.264	0.0151	19
9123.60	[Cl ii]	1F	9122.24	-44.69	1.918	0.110	11
9210.28	He i	6/9	9209.03	-40.69	1.850	0.105	11
9213.20	He i	7/9	9211.81	-45.23	0.708	0.0403	13
9229.01	H i	P9	9227.67	-43.53	44.15	2.515	10	b
9516.57	He i	4/7	9515.35	-38.43	1.272	0.0720	12
9530.60	[S iii]	1F	9529.68	-28.94	1467.4	83.02	8	c
9603.44	He i	2/6	9602.12	-41.21	0.4626	0.0261	15
9824.13	[C i]	3P-1D	9822.00	-65.00	2.619	0.147	11
9850.26	[C i]	3P-1D	9848.20	-62.70	7.668	0.431	11	b
9902.30	[Kr iii]	3P-1D	9901.82	-14.53	3.057	0.172	11	c
9903.46	C ii	17.02	*	*	*	*	*
9982.46	O ii	G[5]0-2[6]	9981.00	-43.55	0.247	0.0139	19
9988.54	O ii	G[5]0-2[6]	9988.68	4.20	0.705	0.0396	13
9990.08	O ii	D[3]0-0[4]	*	*	*	*	*
9991.48	O ii	D[3]0-0[4]	*	*	*	*	*
10008.87	O ii	G[4]0-1[5]	10007.46	-43.13	0.144	0.0081	25
10027.70	He i	6/7	10026.12	-47.24	7.034	0.394	10	b
10031.20	He i	7/7	10029.66	-46.02	1.853	0.104	11
10049.40	H i	P7	10047.92	-44.15	100.0	5.600	10
10072.07	He i	5/7	10070.70	-40.84	0.450	0.0252	14
	?		*		*	*	*	*
10138.40	He i	10/7	10137.04	-40.22	0.524	0.0293	14
10286.70	[S ii]	3F	10285.43	-37.01	31.05	1.732	10	b
10310.70	He i	4/6	10309.85	-24.71	2.640	0.147	11	b
10320.50	[S ii]	3F	10318.90	-46.48	32.32	1.801	10
10336.40	[S ii]	3F	10334.83	-45.54	24.93	1.389	10
10370.50	[S ii]	3F	10368.88	-46.83	10.54	0.587	11
10397.50	[N i]	3F	10396.38	-32.29	8.154	0.454	11	b
10407.40	[N i]	3F	10405.86	-44.36	4.351	0.242	11
a Affected by charge transfer.
b Affected by telluric emission.
c Affected by atmospheric absorption bands.
d Normalized to H$\beta$ until [S ii] 6730 $\AA$ line. Normalized to P7 10047 $\AA$ from C ii 6780 $\AA$ to [N i] 10407 $\AA$.
e Where $I$ is the reddened corrected flux, with c(H$\beta$)=0.63, in units of $100.00=1.032\times {10}^{-11}$ erg cm-2 s-1.
f Colons indicates errors larger than 40 $\%$

Table 15. Table 3: Observed and reddening corrected line ratios (F(H β 𝛽 \beta ) = 100) and line identifications in the FIRE spectrum of NGC 5315.

${\lambda }_0$ (Å )	Ion	Mult.	${\lambda }_{\mathrm{obs}}$	$V_{\mathrm{rad}}$ (km s-1)	$F\left(\lambda \right)/F\left(Href\right)$d	$I\left(\lambda \right)/I\left(H\beta \right)$e	Err ($\%$)f	notes
8316.55	H i	P26	8315.52	-37.07	2.772	0.162	10
8325.71	H i	P25	8324.70	-36.32	3.815	0.223	9
8336.07	H i	P24	8335.18	-31.99	3.937	0.230	9
8347.84	H i	P23	8346.82	-36.95	5.362	0.313	8
8361.30	H i	P22	8361.31	0.27	7.923	0.463	8
8364.04	He i	68	*	*	*	*	*
8376.78	H i	P21	8376.25	-19.13	6.377	0.372	8
8394.71	H i	P20	8394.08	-22.55	5.749	0.335	8
8399.71	He i	6/19	8401.55	65.67	0.378	0.0220	:
8402.13	He i	7/19	*	*	*	*	*
8415.63	H i	P19	8415.01	-22.05	6.473	0.377	8
8424.28	He i	6/18	8423.44	-29.80	0.363	0.0211	30
8440.28	H i	P18	8439.80	-17.15	6.498	0.379	8
8448.57	O i	4	8448.28	-10.33	11.02	0.642	7
8448.68	O i	4	*	*	*	*	*
8449.08	O i	4	*	*	*	*	*
8453.48	He i	6/17	8453.54	2.13	0.271	0.0158	:
8455.93	He i	7/17	*	*	*	*	*
8469.58	H i	P17	8468.99	-20.80	8.952	0.521	8
8483.53	[Cl iii]	3F	8482.63	-31.82	0.789	0.0459	19
8488.60	He i	6/16	8488.16	-15.57	0.676	0.0393	22
8502.50	[Cl iii]	3F	8504.11	56.73	10.59	0.615	8
8504.82	H i	P16	*	*	*	*	*
8531.38	He i	6/15	8531.54	5.80	0.870	0.0505	19
8533.87	He i	7/15	*	*	*	*	*
8547.73	H i	P15	8547.15	-20.27	11.39	0.661	7
8581.05	[Cl ii]	1F	8580.46	-20.61	5.990	0.347	8
8584.22	He i	6/14	8584.73	17.80	1.246	0.0722	15
8586.74	He i	7/14	*	*	*	*	*
8600.75	H i	P14	8600.17	-20.40	12.09	0.701	7
8619.32	[Fe ii]	13F	8618.13	-41.33	2.133	0.123	11
8650.64	He i	6/13	8650.17	-16.32	0.937	0.0542	18
8653.20	He i	7/13	*	*	*	*	*
8667.40	H i	P13	8666.87	-18.16	15.47	0.895	7
8735.83	He i	6/12	8735.30	-18.04	1.242	0.0716	15
8738.44	He i	7/12	*	*	*	*	*
8752.87	H i	P12	8752.47	-13.85	20.14	1.161	7
8779.34	He i	4/9	8778.77	-19.63	1.340	0.0772	14
8831.82	[S iii]	1F	8830.69	-38.58	0.717	0.0413	21
8847.83	He i	6/11	8847.00	-28.12	1.171	0.0673	15
8850.47	He i	7/11	8849.73	-24.96	0.689	0.0396	21
8856.63	He i	5/11	8855.77	-29.11	0.191	0.0110	:
8857.71	[Se iii]	3P-1D	*	*	*	*	*
8865.22	H i	P11	8864.77	-15.48	21.11	1.213	7
8894.37	[Fe ii]	13F	8893.71	-21.56	0.398	0.0228	30
8999.26	He i	6/10	8999.22	-1.33	2.404	0.138	10
9002.21	He i	7/10	*	*	*	*	*
9017.38	H i	P10	9016.76	-20.91	32.13	1.839	6
9071.09	[S iii]	1F	9070.84	-8.18	1029	58.82	6
9126.10	[Cl ii]	1F	9125.24	-28.36	2.126	0.121	11
9212.85	He i	6/9	9212.09	-24.80	2.815	0.160	10
9215.76	He i	7/9	9214.90	-28.04	0.830	0.0473	19
9231.54	H i	P9	9231.02	-16.85	43.96	2.503	6
9466.17	He i	1/5	9464.99	-37.38	6.913	0.392	7
9519.44	He i	4/7	9518.95	-15.49	6.077	0.344	7
9533.21	[S iii]	1F	9532.80	-13.10	2848	161.1	6
9548.59	H i	P8	9547.93	-20.65	68.41	3.869	6
9606.07	He i	2/6	9604.93	-35.83	0.590	0.0333	23
9628.34	He i	1P0-1D	9627.71	-19.50	0.642	0.0362	22
9826.82	[C i]	3P-1D	9826.04	-24.01	1.753	0.0886	12
9852.96	[C i]	3P-1D	9852.25	-21.61	4.796	0.270	7
9905.00	[Kr iii]	3P-1D	9905.53	15.95	3.613	0.203	8
9906.18	C ii	17.02	*	*	*	*	*
9985.19	O ii	G[5]0-2[6]	9984.70	-14.56	0.321	0.0180	35
9992.46	O ii	G[5]0-2[6]	9991.98	-14.43	1.228	0.0688	15
9993.02	O ii	D[3]0-0[4]	*	*	*	*	*
9993.21	O ii	D[3]0-0[4]	*	*	*	*	*
10013.62	O ii	G[4]0-1[5]	10011.96	-49.70	0.706	0.0396	21
10030.50	He i	6/7	10029.93	-17.04	6.531	0.366	7
10033.90	He i	7/7	10033.20	-20.91	2.051	0.115	11
10052.16	H i	P7	10051.51	-19.24	100.0	5.60	5
10074.81	He i	5/7	10073.73	-32.17	0.407	0.0228	29
	?		*	*	*	*	*
10289.52	[S ii]	3F	10288.95	-16.59	19.83	1.105	6
10314.13	He i	4/6	10313.41	-20.81	3.987	0.222	8
10323.33	[S ii]	3F	10322.74	-17.09	25.16	1.402	6
10339.23	[S ii]	3F	10338.16	-31.11	22.72	1.266	6
10373.34	[S ii]	3F	10372.66	-19.71	8.937	0.498	6
10400.35	[N i]	3F	10400.08	-7.77	4.607	0.256	8
10410.25	[N i]	3F	10409.58	-19.36	3.334	0.186	8
10670.61	He i	3P0-3S	10670.13	-13.49	0.611	0.0338	23
10833.31	He i	1/4	10832.78	-14.58	1581	87.39	5	c
10916.03	He i	6/6	10915.88	-4.25	15.73	0.869	6
10920.05	He i	7/6	10919.35	-19.22	4.086	0.226	8
10941.09	H i	P6	10940.45	-17.57	174.9	9.650	5
10993.84	[Se iii]	3P-1D	10991.79	-55.90	0.217	0.0120	:
10999.70	He i	5/6	10998.88	-22.40	0.963	0.0531	17
11016.09	He i	1S-1P0	11015.21	-23.88	0.898	0.0495	17
11048.01	He i	1P0-1D	11047.31	-18.86	1.704	0.0939	12
11290.41	O i	3P-3D0	11288.91	-39.85	3.004	0.165	9
11471.30	[P ii]	3P-1D	11470.11	-31.10	1.802	0.0987	9
	?		11696.42		0.349	0.0191	34
11751.45	C i	3D-3F0	11750.50	-24.24	0.369	0.0201	:
11756.53	C i	3D-3F0	11756.30	-5.87	0.528	0.0288	:
11758.00	C i	3D-3F0	*	*	*	*	*
	?		11864.36		16.68	0.909	6
	?		11870.45		1.055	0.0575	16
11886.10	[P ii]	3P-1D	11885.48	-15.70	7.498	0.409	6
11942.40	C ii	2F0-2G	11940.71	-44.13	1.241	0.0676	14
11972.28	He i	3P0-3D	11971.73	-13.67	6.372	0.347	6
12226.29	[Fe ii]	a6D-a4D	12224.29	-49.41	1.291	0.0701	14	b
12530.93	He i	1/3	12530.25	-16.21	9.343	0.505	6
12570.21	[Fe ii]	a6D-a4D	12569.22	-24.29	7.507	0.406	6
12675.32	[Fe ii]	a4D-a2G	12674.46	-20.26	0.304	0.0164	37
	?		12692.18		7.174	0.387	6
12706.91	[Fe ii]	a6D-a4D	*	*	*	*	*
12788.42	He i	6/5	12787.88	-12.60	23.28	1.256	5
12794.00	He i	7/5	12793.26	-17.34	8.353	0.451	6
12821.59	H i	P5	12821.01	-13.48	344.2	18.57	5
12849.46	He i	3P0-3S	12847.83	-37.99	1.623	0.0875	12
12946.20	[Fe ii]	a6D-a4D	12944.82	-32.66	1.205	0.0649	15
12971.98	He i	1P0-1D	12971.62	-8.26	1.559	0.0840	13
12988.43	He i	5/5	12987.38	-24.28	2.013	0.108	11
13005.69	[Fe ii]	a4D-a2G	13001.23	-102.81	0.306	0.0165	37
13167.49	O i	3P-3S0	13167.56	1.59	3.870	0.208	8
	O i	3P-3S0	*	*	*	*	*
	O i	3P-3S0	*	*	*	*	*
13209.15	[Fe ii]	a6D-a4D	13207.77	-31.36	2.044	0.110	11
13281.40	[Fe ii]	a6D-a4D	13280.21	-26.94	0.714	0.0384	21
13417.97	[Fe ii]	a4D-a2G	13415.37	-58.09	0.385	0.0206	32
14819.71	H i	Br32	14819.41	-6.16	7.192	0.0647	16
14835.15	H i	Br31	14835.19	0.77	3.229	0.0291	24
14852.20	H i	Br30	14851.12	-21.71	8.267	0.0744	15
14871.08	H i	Br29	14870.04	-20.93	8.494	0.0764	15
14892.08	H i	Br28	14891.49	-11.86	8.408	0.0757	15
14915.52	H i	Br27	14914.91	-12.17	5.233	0.0471	19
14941.81	H i	Br26	14941.58	-4.67	8.123	0.0731	15
14971.42	H i	Br25	14970.50	-18.44	11.057	0.0995	12
15004.96	H i	Br24	15004.16	-15.98	10.549	0.0949	12
15043.15	H i	Br23	15042.23	-18.34	10.486	0.0944	12
15078.36	He i	3F0-3D	15077.41	-18.88	3.587	0.0323	22
15086.89	H i	Br22	15086.54	-7.00	17.292	0.156	10
	He i	1S0-1P0	*	*	*	*	*
15137.36	H i	Br21	15136.72	-12.59	12.70	0.114	11
15195.99	H i	Br20	15195.17	-16.21	14.45	0.130	11
15264.71	H i	Br19	15264.82	2.15	25.63	0.231	9	c
15338.94	[Fe ii]	a4F-a4D	15338.65	-5.65	7.758	0.0698	14
15345.98	H i	Br18	15345.41	-11.19	18.02	0.162	10
15443.14	H i	Br17	15442.70	-8.52	20.58	0.185	10
15560.70	H i	Br16	15560.13	-11.00	24.90	0.224	9
15696.27	He i	3F0-3D	15694.61	-31.71	3.998	0.0360	21
15704.95	H i	Br15	15704.40	-10.50	28.06	0.254	9
	?		15808.91		3.152	0.0284	25
15876.18	He i	3F0-3D	15875.02	-21.84	3.596	0.0324	23
15884.88	H i	Br14	15884.50	-7.17	32.27	0.290	9
15999.15	[Fe ii]	a4F-a4D	15997.91	-23.16	4.330	0.0390	20
16113.71	H i	Br13	16112.96	-14.03	39.75	0.358	8	c
16402.94	He i	3F0-3D	16401.61	-24.31	3.369	0.0303	25
16411.67	H i	Br12	16411.20	-8.67	46.06	0.415	8
16440.02	[Fe ii]	a4F-a4D	16438.47	-28.21	22.49	0.202	10	b
	?		16601.43		5.988	0.0539	18
16642.25	[Fe ii]	a4F-a4D	16641.14	-19.99	3.191	0.0287	26
16773.41	[Fe ii]	a4F-a4D	16776.05	47.26	9.247	0.0832	14
	?		*	*	*	*	*
16802.44	He i	3F0-3D	16800.92	-27.10	8.137	0.0732	15
16811.11	H i	Br11	16810.63	-8.60	69.23	0.623	8
17007.64	He i	3P0-3D	17006.67	-17.18	49.52	0.446	8
17356.50	He i	3F0-3G	17355.82	-11.70	10.66	0.0959	14
17366.85	H i	Br10	17365.92	-16.09	100	0.900	8
17454.15	[Fe ii]	a4F-a4D	17454.94	13.55	5.766	0.0513	19
	He i	3S-3P0	*	*	*	*	*
	?		17639.98		10.59	0.0919	14
			19366.14		0.520	0.0145	33
19411.45	He i	3D-3F0	19409.76	-26.10	0.883	0.0247	23
19418.88	He i	1D-1F0	19416.10	-42.92	1.379	0.0386	17
19438.91	He i	3F0-3G	19438.54	-5.71	3.292	0.0922	10
19450.87	H i	Br8	19449.19	-25.88	38.53	1.079	5
19548.63	He i	3D-3P0	19546.85	-27.23	12.31	0.345	6
20430.41	He i	3S-3P0	20429.07	-19.68	1.148	0.0321	19
20586.90	He i	1S-1P0	20586.02	-12.88	73.02	2.045	5
20607.36	He i	3P0-3D	20605.64	-25.00	1.552	0.0435	16
21125.89	He i	3P0-3S	21124.68	-17.10	5.330	0.149	8
21137.80	He i	1P0-1S	21136.21	-22.53	1.809	0.0507	14
21209.90	H2	1-0 S(1)	21213.18	46.36	1.181	0.0331	19
21613.69	He i	3D-3F0	21612.37	-18.31	3.788	0.106	9
21622.91	He i	1D-1F0	21621.48	-19.76	1.327	0.0372	17
21647.45	He i	3F0-3G	21646.63	-11.39	13.45	0.377	6
21661.20	H i	Br7	21660.35	-11.78	100	2.800	5
21986.00	[Kr iii]	3P-3P	21985.13	-11.86	1.986	0.0556	14
22186.75	[Fe iii]	3H-3G	22185.58	-15.86	0.740	0.0207	26
22864.00	[Se iv]	–	22860.22	-49.56	2.952	0.0827	11
23066.55	H i	Pf46	23068.02	19.17	0.303	0.0085	:
	?		*	*	*	*	*
23078.95	H i	Pf46	23078.65	-3.83	0.261	0.0073	:
	?		*	*	*	*	*
23092.22	H i	Pf44	23092.20	-0.19	0.191	0.0053	:
23106.44	H i	Pf43	23105.93	-6.59	0.282	0.0079	:
23121.71	H i	Pf42	23120.41	-16.83	0.311	0.0087	:
23138.13	H i	Pf41	23136.73	-18.17	0.396	0.0111	39
	He i	3D-3F0	*	*	*	*	*
23151.23	He i	3P0-3D	23153.83	33.72	0.465	0.0130	35
23155.83	H i	Pf40	*	*	*	*	*
23161.39	He i	3D-3F0	*	*	*	*	*
23174.94	H i	Pf39	23172.04	-37.46	0.873	0.0244	23
23195.61	H i	Pf38	23192.57	-39.23	0.664	0.0186	28
23218.02	H i	Pf37	23215.04	-38.42	0.894	0.0250	23
23242.37	H i	Pf36	23238.49	-50.02	0.809	0.0226	24
23268.90	H i	Pf35	23264.42	-57.66	0.802	0.0225	24
23297.87	H i	Pf34	23292.79	-65.32	0.927	0.0259	22
23329.59	H i	Pf33	23324.07	-70.99	1.031	0.0289	21
23364.44	H i	Pf32	23358.47	-76.61	1.193	0.0334	19
23402.83	H i	Pf31	23396.35	-83.06	1.151	0.0322	19
23445.28	H i	Pf30	23437.89	-94.46	1.377	0.0386	17
23492.37	H i	Pf29	23483.66	-111.10	1.957	0.0548	14
23544.81	H i	Pf28	23537.81	-89.14	1.522	0.0426	16
23603.47	H i	Pf27	23596.44	-89.25	1.621	0.0454	15
23669.37	H i	Pf26	23662.81	-83.03	1.942	0.0544	14
23730.67	He i	3F0-3D	23724.99	-71.81	0.259	0.0073	:
23743.77	H i	Pf25	23737.93	-73.73	1.924	0.0539	14
23828.23	H i	Pf24	23824.48	-47.19	2.534	0.0709	12
23924.68	H i	Pf23	23922.53	-26.88	2.507	0.0702	12
24035.52	H i	Pf22	24037.76	27.90	2.669	0.0747	11
24065.96	H2	1-0 Q(1)	24069.60	45.34	0.862	0.0241	23
24163.85	H i	Pf21	24173.09	114.61	2.337	0.0654	12
24243.61	H2	1-0 Q(3)	24245.20	19.61	1.544	0.0432	16
24313.62	H i	Pf20	24331.63	222.07	3.688	0.103	10
	?		24493.32		1.019	0.0285	21
24490.00	H i	Pf19	24522.68	400.02	4.336	0.121	9
24699.87	H i	Pf18	24753.25	647.86	5.309	0.149	8
a Affected by charge transfer.
b Affected by telluric emission.
c Affected by atmospheric absorption bands.
d Where $F$ is the unreddened flux measured against the reference line
e Where $I$ is the reddened corrected flux, in units of $100.00=1.032\times {10}^{-11}$ erg cm-2 s-1.
f Colons indicates errors larger than 40 $\%$

Equations

$$[\mathrm{Kr}/\mathrm{Ar}]=(0.472\pm 0.093)+(1.240\pm 0.490)\log (\mathrm{C}/\mathrm{O}).$$

Full text

††pubyear: 2016††pagerange: Neutron-capture element abundances in the planetary nebula NGC 5315 from deep optical and near-infrared spectrophotometry††thanks: This paper includes observations collected at the European Southern Observatory, Chile, proposal number ESO 092.D-0189(A).††thanks: This paper includes data gathered with the 6.5 meter Magellan Telescopes located at Las Campanas Observatory, Chile.–Neutron-capture element abundances in the planetary nebula NGC 5315 from deep optical and near-infrared spectrophotometry††thanks: This paper includes observations collected at the European Southern Observatory, Chile, proposal number ESO 092.D-0189(A).††thanks: This paper includes data gathered with the 6.5 meter Magellan Telescopes located at Las Campanas Observatory, Chile.

1 Introduction

Trans-iron (n-capture) elements are produced in asymptotic giant branch (AGB) stars (1–8 M*⊙) in the intershell region between the H- and He-burning shells, in the so-called s-process (slow neutron-capture process). In these layers neutrons are released by $\alpha$-captures onto 13C (or 22Ne in AGB stars with mass $>3-4$ M⊙*). During the thermally-pulsing phase, convective dredge-up conveys to the stellar surface C and s-process-enriched material, which is then expelled to the interstellar medium in the planetary nebula (hereafter PN) phase and eventually incorporated in a new generation of stars. Thus, the abundances of n-capture elements and the s-process element-by-element enrichment pattern reveal critical information on physical conditions in stellar interiors and the nucleosynthetic histories of stellar populations (Cristallo et al., 2011, 2015; Karakas & Lattanzio, 2014; Trippella et al., 2014, 2016; Ventura et al., 2015).

While s-process nucleosynthesis has historically been studied through stellar spectra, nebular spectroscopy presents unique advantages. PNe allow for the first observational analysis of the lightest n-capture elements (Ge, Se, and Br) and the noble gases Kr and Xe in one of their sites of origin. A crucial difference from AGB stars is that PNe represent the final envelope abundances of their progenitor stars, after the cessation of nucleosynthesis and dredge-up that occurred during the AGB phase. PN abundances are therefore highly valuable for stellar yield determinations. Moreover, nebular abundance determinations provide key constraints to poorly-understood processes in models of stellar evolution and nucleosynthesis, such as the efficiency of third dredge-up (TDU) at low envelope masses, the number of thermal pulses, and treatments of mass loss and convective overshoot (Cristallo et al., 2015; Karakas et al., 2009, 2012; Karakas & Lattanzio, 2014; Trippella et al., 2016; Ventura et al., 2015, and references therein). The number of n-capture elements that can be detected in individual PNe allows for meaningful comparisons with various sets of evolutionary models (e. g., Sterling et al., 2016, hereafter S16). In particular, the optical region is home to numerous n-capture element transitions, including lines from multiple ions of Se, Br, Kr, Rb, and Xe that have been detected in various PNe (e. g. Sharpee et al., 2007; García-Rojas et al., 2012; García-Rojas et al., 2015).

Since the pioneering work by Péquignot & Baluteau (1994), only a few detailed studies of n-capture element abundances based on deep, high-resolution optical spectra have been conducted in PNe. Sharpee et al. (2007) identified lines of Br, Kr, Rb, Xe, Ba, and possibly Te and I in 4 PNe, with 4- and 6-m class telescopes. But at their resolution of $\sim$22,000, many features were not unambiguously detected. García-Rojas et al. (2015) made a detailed n-capture element abundance analysis in NGC 3918, a PN with a high ionization degree and with a C/O ratio close to 1, detecting several ions of Kr, Xe, Rb and Se. This allowed them to test the new ionization correction factors (ICFs) for n-capture elements provided by Sterling et al. (2015) and to compute total abundances with unprecedented accuracy. In contrast, near-infrared (hereafter NIR) lines of n-capture elements have been studied in a large number of PNe. For example, Sterling & Dinerstein (2008)’s survey of [Kr iii] and [Se iv] in 120 PNe resulted in the first overview of s-process enrichments in PNe. However, their abundance determinations were uncertain by factors of 2–3, since only one ion of each element was detected, leading to large and uncertain corrections for unobserved ions.

In this work we combine optical and NIR spectroscopy to study NGC 5315, a PN that has been classified as an almost spherical (slightly elliptical) PN with a complicated structure, including a somewhat broken ring. The H-deficient WC4 central star has a temperature of about 76-79 kK (Marcolino et al., 2007; Todt et al., 2015), which is considerably lower than other early-type [WR] stars (120$-$150 kK) . Several studies have focused on the chemical content in NGC 5315, but draw different conclusions, especially regarding the N/O ratio which is important for constraining the progenitor mass. Pottasch et al. (2002) combined IUE ultraviolet and ISO infrared spectra with optical data to investigate chemical abundances, and found a relatively high N/O$\sim$0.88. Milingo et al. (2010) found a very similar value from optical/NIR spectrophotometry, as did Peimbert et al. (2004). The excess of He and N led these authors to classify NGC 5315 as a Peimbert Type I PN (He/H $>0.125$ and N/O $>0.5$, or N/O $>0.8$ Peimbert, 1978; Kingsburgh & Barlow, 1994), although it is not an extreme object. The high concentration of He and N may be explained with the occurrence of the second dredge-up and CN-cycling during hot bottom burning (HBB), which are activated in AGB stars with M $>$ 3–4 M*⊙* (Becker & Iben, 1979; Boothroyd et al., 1993; , Di Criscienzo et al.2016). Karakas & Lugaro (2016) found through detailed theoretical models that HBB activation requires a minimum progenitor mass between 4 and 5 M*⊙*.

However, other investigations of NGC 5315 (de Freitas Pacheco et al., 1991; Samland et al., 1992; Tsamis et al., 2003; Dufour et al., 2015) do not find Type I abundances, calling into question whether the PN indeed derives from a more massive progenitor. It should be noted that only N+ has collisionally-excited optical transitions. Since this is a trace ion, the ionization correction factor (ICF) for N can be large and uncertain when only optical data are used in deriving its abundance. The detection of N iii$]$ and N iv$]$ lines in UV spectra can lead to much more accurate N abundance determinations. Both Tsamis et al. (2003) and Dufour et al. (2015) utilize optical and UV data, finding N/O ratios of 0.54 and 0.41, respectively, contrasting with the much higher value of Pottasch et al. (2002). This variance may illustrate the uncertainties of abundances computed with highly temperature-sensitive UV lines. As for the C/O abundance ratio, the values found in the literature range from 0.35 (Tsamis et al., 2003) to 0.95 (Peimbert et al., 2004). The N/O and C/O ratios are very useful for constraining the progenitor mass, and the disparity in the ratios found, even from UV observations that provide access to a wide range of C and N ions, highlight the enigmatic nature of NGC 5315’s progenitor. The different emission lines and methods used to derive C/O and N/O ratios will be discussed in detail in Sect. 7.

Abundance determinations of n-capture elements in PNe can provide more restrictive constraints to the mass of the progenitor star. Observations of n-capture elements in Type I PNe suggest that they exhibit little if any s-process enrichments (Sterling & Dinerstein, 2008; Sterling et al., 2015), although in such studies only 1 or 2 n-capture elements were studied. On the other hand, very low massive progenitors ($<1.5$ M*⊙*) will also show little s-process enrichments as they have not gone through the third dredge up. However Rb enrichments in PNe can help to disentangle this puzzle, because it is an indicator of the main neutron source during the AGB phase (13C or 22Ne) and, hence of the mass of the progenitor star (Busso et al., 1999).

There are relatively few observational constraints on nucleosynthetic models for AGB stars with M $>$ 3–4 M*⊙. García-Hernández et al. (2006) and García-Hernández et al. (2009) reported large Rb enrichments in 4–8 M⊙* Galactic and Magellanic cloud AGB stars, respectively. García-Hernández et al. (2007) and García-Hernández et al. (2013) provided Zr and Rb abundances, respectively, for massive galactic AGB stars; however, they found extremely high Rb/Zr ratios that were not predicted by nucleosynthesis models. Zamora et al. (2014) took into account circumstellar effects on AGB stars to correct Rb abundance computations to Rb/Zr values in better agreement with results from AGB nucleosynthesis models for stars with masses between 4–8 M*⊙* (-0.2 $\leq$ [Rb/Zr] $<$ 0.6) (Karakas et al., 2012). According to the models of Karakas et al. (2012), Kr and Se should be enriched in the objects which show Rb enrichment for the 22Ne neutron source. Therefore, deep optical and NIR spectroscopy of PNe, with the aim of detecting emission lines of Rb, Kr and Se of different ionization stages, is a very valuable tool to improve the accuracy of these n-capture element abundance determinations and hence, better constrain the initial mass of the progenitor stars of these objects.

This study is a continuation of our work to collect deep, high-resolution spectra of PNe to investigate objects covering different ionization degrees to detect as many ions as possible of n-capture elements in order to improve the accuracy of their abundance determinations. These observations aim to address scientific goals including: i) to study the correlation between different n-capture elements and C enrichment, predicted by models; ii) to study the correlation between the pattern of n-capture element abundances and the mass of the progenitor star, which is modulated by the nuclear reaction activated in each mass range (Karakas et al., 2012; van Raai et al., 2012); iii) to use detections of multiple ions of individual n-capture elements to test the atomic data and ICF prescriptions (Sterling et al., 2015) for these species. Our first results for NGC 3918 have been published by García-Rojas et al. (2015).

The observations and data reduction are described in Sect. 2. The identification of lines and reddening correction is presented in Sect. 3. In Sect. 4 we compute the physical conditions. In Sects. 5 and 6 we compute ionic and total abundances. In Sect. 7 we discuss the results and in Sect. 8 we draw some conclusions.

2 Observations and data reduction

2.1 UVES

The optical spectra of NGC 5315 were taken with the Ultraviolet-Visual Echelle Spectrograph (D’Odorico et al., 2000), attached to the 8.2m Kueyen (UT2) Very Large Telescope at Cerro Paranal Observatory (Chile) in service mode. The observations were performed during 3 nights under clear/dark conditions and the seeing remained below 1.5*′′* during the whole run (see Table 1) .

The slit width was set to 1 arcsec, which provides an effective spectral resolution R$\sim$ 40000. See García-Rojas et al. (2015) for further details on the instrument setup. The journal of observations is shown in Table 1. The atmospheric dispersion corrector (ADC) was used to compensate for atmospheric dispersion. The UVES spectra are divided in four spectral ranges (B1, B2, R1 and R2). We took individual exposures of 630 s in each configuration and we followed the same sequence as in García-Rojas et al. (2015). After combining all the extracted spectra we obtained a total exposure time of 2.10 h in each configuration. Additional single short exposures of 60 s during the first night were taken to obtain non-saturated flux measurements for the brightest emission lines. Reduction of the raw frames was done in the same way as in García-Rojas et al. (2015) for each night. The extracted spectra cover an area of 3.95*′′* $\times$ 1*′′* common to all spectral ranges. Spectrophotometric standards Feige 67 (Oke, 1990) and LTT 3218 (Hamuy et al., 1992, 1994) were observed to perform the flux calibration and were also fully reduced with the pipeline. We used iraf111IRAF is distributed by the National Optical Astronomy Observatories, which are operated by the Association of Universities for Research in Astronomy, Inc., under cooperative agreement with the National Science Foundation. (Tody, 1993) to perform the flux calibration and the radial velocity corrections. Flux-calibrated, radial velocity corrected, one-dimensional spectra for each night were finally co-added to obtain the final one-dimensional optical spectra analyses in this paper.

In Figure 1 we show a high spatial resolution H$\alpha$ image of NGC 5315 from the HST archive. The slit center was set 1.0*′′* east to the central star of NGC 5315 oriented N-S (PA=0*∘*), covering the brightest area of NGC 5315.

2.2 FIRE

The NIR spectra of NGC 5315 were taken with the Folded-port InfraRed Echellette (FIRE, Simcoe, 2013) spectrograph attached to the 6.5m Magellan Baade Telescope (MBT) located at Las Campanas Observatory in Chile. The observations were performed on 2013 August 13 and the seeing oscillated between 1.8*′′–2.9′′* during the run.

The slit was located across the central star (PA=90*∘) (see Figure 1) is 7′′* long with a width of 0.75*′′, which lead to a resolution of R$\sim$ 4800. The spectral range is 0.8 – 2.5 $\micron$, which covers J, H and K bands. Wavelength calibrations were performed in vacuum using a Th-Ar lamp, instead of the air wavelengths used in the UVES observations. Unfortunately, the lack of lines in the Th-Ar lamps at wavelengths larger than 2.3 $\mu$m led to a poor wavelength solution for the range 2.3-2.5 $\mu$m; however as this range is dominated by the Pfund H i series, this did not affect the identification of the emission lines. For flux calibration and telluric correction, A0V standard stars were observed. We took 20 exposures of 30s, which led to a total integration time of 600s. We obtained 20 sky frames nodded 60′′* along the slit direction, in order to subtract telluric features from the final spectra. Data reduction was performed with the FIRE reduction software package, FIREHOSE, using optimal extraction. This pipeline is based on the MASE pipeline for the MagE spectrograph (Bochanski et al., 2009), and a brief description of its performance can be found in Simcoe (2013). The journal of observations is shown in Table 1.

3 Line fluxes, line identification and extinction correction

The line fluxes were measured with the splot routine of the iraf package. As in García-Rojas et al. (2015), we decided to integrate over the entire line profile as set between two limits over a local continuum estimated by eye, as opposed to fitting an analytic function (such as a Gaussian) to the line profile. The UVES spectra present many misleading features such as telluric lines and internal reflections caused by the dichroic. For this reason, we had to take care to correctly distinguish nebular emission lines. Telluric emission lines are easily recognized and discarded, thanks to their peculiar shape in the 2D spectra (narrow profile and homogeneous emission along the slit). For the weakest telluric features we used two catalogues: Hanuschik (2003) in the optical and Oliva et al. (2015) in the NIR. When emission line intensities were affected by internal reflections or telluric absorption/emission, we include a note of caution in the line identification tables (Tables LABEL:lineid_uves and LABEL:lineid_fire).

The UVES spectrum covers four spectral ranges, which overlap at the edges. In order to create a homogeneous set of data, we used the H9 $\lambda$3835 line, which lies in the overlapping region between the two spectral ranges B1 and B2, to normalize the line intensities, which were later re-scaled to H$\beta$. Lines in the B2 and R1 ranges were directly scaled to H$\beta$ which is present in both ranges. [S ii] $\lambda$6716 and $\lambda$6730 $\AA$ are the only lines common to the R1 and R2 settings. Unfortunately, in the R2 range, they lie at the extreme blue edge of the spectrum and their fluxes are not reliable. Therefore, we could not re-scale the fluxes in R2 to H$\beta$ and we normalized all the lines to P7 10047 $\AA$. The NIR FIRE spectrum does not include the H$\beta$ line, thus we used a different H i line in each three spectral bands to normalize the line fluxes: P7 10047 $\AA$ for the J-band, Br10 1.7367 $\mu$m for the H-band and Br$\gamma$ 2.1661 $\mu$m for the K-band.

For the reddening correction, we assumed the extinction law of Cardelli et al. (1989) with Rv=3.1. Using Balmer lines from UVES spectra we found a value of c(H$\beta$) = 0.63 $\pm$ 0.02. The deepest study of this object in the optical range until now is that of Peimbert et al. (2004), which find a value of c(H$\beta$) = 0.74 $\pm$ 0.04; while other results from optical data give values in better agreement with our calculation: Dufour et al. (2015) find c(H$\beta$) = 0.56 from STIS/HST data and Cahn et al. (1992) compute c(H$\beta$) = 0.60. Since a normalization to H$\beta$ is not available for the spectral range R2 of UVES and for the J-band of the FIRE spectrum, we considered the extinction coefficient relative to P7 10047 $\AA$, c(P7)=0.05 $\pm$ 0.01, calculated assuming the extinction law by Cardelli et al. (1989) with Rv=3.1. Finally, for the H-band and the K-band we used the theoretical ratios between the intensities of the reference lines used in each band (Br10 for the H-band and Br$\gamma$ for the K-band) and H$\beta$, provided by Storey & Hummer (1995) with $n_{\rm e}$=30,000 cm*-3* and $T_{\rm e}$=10,000 K, in order to re-scale the final intensities to H$\beta$.

We have detected about 700 emission lines. Many are permitted lines of H i, He i, O i, O ii, N ii, C i, C ii, S iii, Ne i, Ne ii, Mg ii and Si ii. We also detect several forbidden and semi-forbidden lines from ions such as [N i], [N ii], [O i], [O ii], [O iii], [Ne iii], Mg i], [P ii], [S ii], [S iii], [Cl ii], [Cl iii], [Cl iv], [Ar iii], [Ar iv], [Cr ii], [Cr iii], [Mn iii], [Fe ii], [Fe iii], [Kr iii], [Kr iv], [Se iii], [Se iv] and possibly [Xe iv] and [Br iii]. In the NIR spectrum, we also detect some molecular H2 1–0 transitions, but there are no signs of vibrationally excited H2 lines. The depth of our spectra allows us to detect lines as faint as 10*-5*$\times$I(H$\beta$) and 10*-3*$\times$I(Br$\gamma$) in the optical and NIR spectra, respectively. The identifications and adopted laboratory wavelengths of the lines are based on several previous identifications in the literature (e. g. García-Rojas et al., 2015; Hora & Latter & Deutsch, 1999; Peimbert et al., 2004; Rudy et al., 2001, and references therein). We also made use of Peter van Hoof’s atomic line list v2.05B18222http://www.pa.uky.edu/$\sim$peter/newpage/. Details on the identification of n-capture element emission lines are given in Sect. 3.1.

To determine the line flux uncertainties, we considered individually each spectral range (B1, B2, R1, and R2) for UVES and the entire FIRE spectrum. Several lines were chosen in each of these ranges covering the whole range of measured fluxes, i. e. 10${}^{-5}\leq$ $F$($\lambda$)/$F$(H$\beta$) $\leq$10; uncertainties in individual fluxes were determined by choosing the highest and lowest reasonable values for the continuum by about 1$\sigma$ for the selected lines. To assign an error to the whole set of emission lines we made a logarithmic interpolation between relative intensities and measured uncertainties. An error of 5$\%$ for the flux calibration was added quadratically to all line flux uncertainties. In the final intensity uncertainties we also took in account the error for the extinction correction. Line intensities and identifications are presented in Table LABEL:lineid_uves for the UVES spectrum and in Table LABEL:lineid_fire for the FIRE spectrum.

In order to deblend lines with almost coincident central wavelengths, we computed a model using Cloudy v.13.03 (Ferland et al., 2013) to predicted line fluxes. Stellar temperature and luminosity, nebular diameter and elemental abundances of common elements were optimized to reproduce observed lines in the optical/NIR range from our data and select UV/IR lines from the Pottasch et al. (2002) spectra. The parameters obtained are very similar to those shown in Table 2 of Sterling et al. (2015). The model considers the whole PN, since we did not attempt to simulate an observation with a slit located on the PN. We believe this is a reasonable assumption owing to the fact that our slits cover all the ionization zones of the nebula. In Table 4 we compare different diagnostic ratios predicted by our photoionization model to the observed ones. In general the agreement between observed and predicted ratios is very good.

3.1 Identification of n-capture ion lines

As n-capture elements have very low abundances (Asplund et al., 2009), their lines are very weak. This is why their detection in nebulae has always been a difficult task. However, since the pioneering work by Péquignot & Baluteau (1994), this field has grown significantly. Several n-capture element lines (particularly those of Kr, Xe and Se) have been detected in Galactic PNe and H ii regions, both in the optical (e. g. García-Rojas et al., 2015, and references therein) and in the NIR (e. g. Dinerstein, 2001; Sterling & Dinerstein, 2008; Blum & McGregor, 2008), and even in other galaxies (Vanzi et al., 2008; Mashburn et al., 2016).

We identify lines of the n-capture element ions [Kr iii], [Kr iv], [Se iii], [Se iv] and possibly [Xe iv] and [Br iii] in NGC 5315 (see Figure 2). The measured fluxes are shown in Table 5. Below, we give details on the identification of these lines in the UVES and FIRE spectra.

We identify lines of two ions of Kr. We detect the [Kr iii] 2.1986 $\mu$m line in the NIR spectrum. This line and the [Se iv] 2.2864 $\micron$ line were first identified by Dinerstein (2001), and are the most widely detected n-capture emission lines in PNe (Sterling & Dinerstein, 2008). We also detect the faint [Kr iii] $\lambda$6826.70 line in our optical UVES spectrum. However, although our optical spectrum is of very high resolution, the velocity field of NGC 5315 widens emission lines, and the flux of this line can be affected by the faint He i 3s 3S–16p 3P0 $\lambda$6827.88 line (Péquignot & Baluteau, 1994). We used the theoretical ratio between He i lines of the same series (specifically, He i $\lambda\lambda$6934, 7161 and 7816) using our Cloudy model (see previous section) to estimate that this He i line contributes 26% of the total measured flux. The Kr2+ abundance calculated with the corrected flux is in a good agreement with that from the [Kr iii] 2.1986 $\mu$m line (see Section 5.1). [Kr iii] $\lambda$9902.30 also falls in the spectral range of UVES and FIRE but is strongly blended with the relatively bright C ii $\lambda$9903.46 line. We find that the ratio [Kr iii] 9902.30/6826.70 = 0.078 for the assumed physical conditions (see Sect. 4), and hence [Kr iii] $\lambda$9902.30 accounts for only 1% of the total measured flux of the line. Therefore we do not consider this [Kr iii] line to be detected. We also detect two [Kr iv] lines, $\lambda\lambda$5346.02 and 5867.74, which are the brightest n-capture lines detectable in the optical spectra of PNe (e. g. García-Rojas et al., 2015). Given the relatively high brightness of these lines and the consistency between the computed abundances, we consider them well identified.

Thanks to the wide spectral range covered by our spectra, we detect three lines of Se ions. [Se iii] $\lambda$8855.28 is detected in both UVES and FIRE spectra. It has almost the same wavelength as a weak He I $\lambda$8854.20 line, making its detection a delicate matter (García-Rojas et al., 2015). We used the theoretical ratios, computed with our Cloudy model, of He i lines belonging to the same series as He i $\lambda$8854.20, to correct for the contribution of He i. We find a contribution of 56$\%$ of [Se iii] emission in the FIRE spectra and 53$\%$ in the UVES spectra, and the abundances calculated from both spectra are in good agreement (see Sect. 5.1). Nevertheless, we note that the line detected at 8855 $\AA$ is very faint and therefore the correction made with our photoionization model is very uncertain. Fortunately, we detect another line of Se2+ at 10992 $\AA$ in the FIRE spectrum, which is isolated and was recently identified in NGC 5315 by Sterling et al. (2017). We also detect the [Se iv] 2.2864 $\mu$m line.

We possibly detect [Xe iv] $\lambda$7535.40 line in our UVES optical spectrum. Unfortunately, the [Xe iv] $\lambda$5709.21 line, which comes from the same upper level as [Xe iv] $\lambda$7535.40, could not be observed since it is strongly blended with the permitted N ii $\lambda$5710.77 $\AA$ line. We computed [Xe iv] $\lambda$5709.21/$\lambda$7535.40 = 0.82 for the assumed $T_{\rm e}$ and $n_{\rm e}$ (see Sect. 4). This corresponds to just 8% of the flux measured for the feature at 5710 $\AA$ that we attribute to N ii $\lambda$5710.77 $\AA$, and thus we conclude that [Xe iv] $\lambda$5709.21 line is not detected in our spectrum. Despite the non-detection of other Xe features in the spectrum of NGC 5315, there are no reasonable alternative identifications to the 7535.40 line, and hence we identify it as [Xe iv]. On the other hand, owing to the ionization degree of NGC 5315, Xe2+ should be more abundant than Xe3+. Therefore, we expect to detect the [Xe iii] $\lambda$5846.77 line, which would strengthen the identification of the [Xe iv] $\lambda$7535.40 line. Unfortunately, the [Xe iii] $\lambda$5846.77 line lies on a continuum bump due to the extremely wide stellar C iv $\lambda$5808 line (the red bump of a [WC]-type star) which hampers any detection of faint lines.

We also detect a feature that can be identified as [Br iii] $\lambda$6555.56. However, we do not detect [Br iii] $\lambda$6130.40 and therefore we cannot claim this as a unambiguous detection.

We do not detect any emission line from Rb ions. We estimate an upper limit flux of [Rb iv] $\lambda$5759.55 line using a 3$\sigma$ criterion. We assume that the S/N in this region is too low to detect this line. Therefore, this estimation represents the maximum flux expected at that wavelength taking in account the continuum noise in that region. We also tried to estimate an upper limit to the [Rb iv] $\lambda$1.5973 $\mu$m line, but the telluric correction around this line was unreliable and an estimation of an upper limit is not possible.

4 Physical conditions

Physical conditions were computed using flux ratios of ions which are sensitive to electron temperature and/or to electron density. The computations were carried out with PyNeb v1.0.26 (Luridiana et al., 2015), using the atomic data presented in Table 6. The electron density $n_{\rm e}$ and temperature $T_{\rm e}$ were estimated following the methodology described in previous works of our group (e. g. García-Rojas et al., 2015). As in Peimbert et al. (2004), we initially assumed three different zones in NGC 5315, characterised by low (ionization potential IP $<$ 17 eV), medium (17 eV $<$ IP $<$ 39 eV) and high (IP $>$ 39 eV) ionization. However, given the relatively low excitation of NGC 5315, and the similarities between different diagnostics, we adopted a two-zone model for $n_{\rm e}$ (low-medium and high ionization zones) and $T_{\rm e}$ (low and medium-high ionization zones) (see Table  7).

It is well known that the auroral [N ii] $\lambda$5755 line is affected by recombination. We use Equation 1 of Liu et al. (2000) to compute the recombination contribution so that this line can be used for the $T_{\rm e}$ computation. Since permitted N ii lines can be affected by fluorescence effects in relatively low ionization PNe (see Escalante et al., 2012), we made a preliminary computation of the N2+/H+ ratio by assuming N2+/O2+ (recombination lines) $\approx$ N*+/O+* (collisionally excited lines). We find a recombination contribution os less than 5% of the $\lambda$5755 flux, and therefore neglect it. We computed $T_{\rm e}$ with the diagnostic [S iii] 6312/9069 using [S iii] $\lambda$6312 from UVES and [S iii] $\lambda$9069 from FIRE data. Unfortunately, the [S iii] $\lambda$9069 line in the UVES spectrum is contaminated by atmospheric absorption and cannot be used. Moreover, the [S iii] $\lambda$9530 line in the UVES spectrum is saturated and its measured flux is unreliable. As can be seen in Table 7, $T_{\rm e}$([S iii]) is in good agreement with $T_{\rm e}$([O iii]) and $T_{\rm e}$([Ar iii]), calculated using lines from UVES data. A diagnostic diagram for NGC 5315 is shown in Figure  3.

Given the high density of NGC 5315, the low ionization [O ii] $\lambda$3726/$\lambda$3729 and [S ii] $\lambda$6731/$\lambda$6716 $n_{\rm e}$ diagnostics are saturated, as it is clearly visible from the huge uncertainties of these ratios in Table 7. Therefore, we decided to use the density found with the [Cl iii] $\lambda$5538/$\lambda$5518 diagnostic as representative of the low-medium ionization zone.

We also calculated $n_{\rm e}$ using [Fe ii] line ratios from our FIRE data, which are useful diagnostics for a large range of densities. Using PyNeb and the atomic dataset of Fe+ by Bautista et al. (2015), we derive an electron density of $30000\pm 7000$ cm*-3* which is the average of the values from the [Fe ii] 1.2946 $\mu$m/1.2570 $\mu$m, 1.3281 $\mu$m/1.2570 $\mu$m, 1.5339 $\mu$m/1.2570 $\mu$m and 1.5999 $\mu$m/1.2570 $\mu$m diagnostics. This determination is consistent with the optical density diagnostics of the low-medium ionization zone (see Table 7).

The uncertainties in the physical conditions were calculated using Monte Carlo simulations. For each line flux involved in these calculations, we generated 10000 random values with a Gaussian distribution centered on the observed value with a sigma corresponding to the line flux error. The physical conditions with their uncertainties are presented in Table 7.

5 Ionic abundances

5.1 Collisionally excited lines

Ionic abundances were calculated using collisionally excited lines (CELs), for N, O, Ne, S, Cl, Ar, P, Fe, Se, Br, Kr and Xe ions. Based on the IP of the ions, we choose physical conditions of low ionization (IP $<$ 17 eV) for N*+, O+, P+, S+, Cl+, Fe+, and Fe2+; medium ionization (17 eV $<$ IP $<$ 39 eV) for O2+, S2+, Cl2+, Cl3+, Ar2+, Se2+, Se3+, Br2+* and Kr3+; and high ionization (IP $>$ 39 eV) for Ne2+ and Ar2+. Computations were carried out using PyNeb and the atomic data used are presented in Table 6. Uncertainties in the line intensities and physical conditions were propagated via Monte Carlo simulations. Ionic abundances from CELs are presented in Table 8.

To investigate the extent to which the different slit positions used for the UVES and FIRE observations affect our results, we compare the abundances of ions detected in both data sets (Cl*+, Cl2+, P+, Fe+, Se2+, and Kr2+). As can be seen in Table 9, our optical ionic fractions from UVES spectrum are similar to both the FIRE and the Peimbert et al. (2004) data (our NIR observations were performed using a slit position similar to Peimbert et al. (2004)). In order to be consistent with our results, for this comparison we re-computed the ionic abundances from Peimbert et al. (2004) spectra, using our atomic data set. In Table 8 we can check that the abundances obtained from optical and NIR lines of Cl+, Cl2+, P+, and Fe2+* are in generally good agreement within uncertainties.

For n-capture element ions, we emphasize the good agreement between the abundances of Kr2+ obtained from lines measured in the optical and in the NIR ranges, as well as the abundances obtained using the two available Kr3+ lines in the optical. We also find good agreement between the Se2+/H*+* abundances derived from [Se iii] $\lambda$8855 in the UVES and FIRE spectra (12+log(Se2+/H+)=3.11 and 3.16, respectively). The Se2+ abundance derived from this line is $\sim$0.2 dex lower than that computed from the [Se iii] $\lambda$1.0992 $\mu$m line, which we attribute to the uncertain correction for the contribution of He ito the $\lambda$8854.20 flux. We therefore choose to use the Se2+ abundance obtained from the [Se iii] $\lambda$1.0992 $\mu$m line in our analysis.

We also report the abundances of other 3 n-capture element ions. We calculate the abundance of Br2+ using the [Br iii] $\lambda$6556 line and collision strengths scaled from Kr3+ (see Table 6). The abundance of Xe3+ is calculated using the flux measured for the very faint [Xe iv] $\lambda$7535.40 line. Finally, Rb3+ is calculated from an upper limit flux estimation of the [Rb iv] $\lambda$5759.55 line.

The FIRE NIR spectrum exhibits multiple [Fe ii] lines, which can be used as density diagnostics since they are not prone to fluorescent effects that contaminate optical [Fe ii] lines. The ratio of two lines arising from the same upper level is given by ratio of transition probabilities. So, the expected [Fe ii] 1.2570 $\mu$m/1.6442 $\mu$m, 1.3209 $\mu$m/1.6442 $\mu$m, 1.3209 $\mu$m/1.2570 $\mu$m, 1.2946 $\mu$m/1.3281 $\mu$m, and 1.2946 $\mu$m/1.5339 $\mu$m line ratios are 1.30, 0.35, 0.26, 1.77 and 1.00, respectively, using the transition probabilities of Bautista et al. (2015). The observed line ratios appear to be consistent, within the uncertainties, with the theoretical ones. In particular, the [Fe ii] 1.3209 $\mu$m/1.2570 $\mu$m, 1.2946 $\mu$m/1.3281 $\mu$m, and 1.2946 $\mu$m/1.5339 $\mu$m line ratios are 0.27$\pm$0.04, 1.70$\pm$0.40, and 0.83$\pm$0.16, respectively, which are in good agreement with theoretical expectations, whereas the [Fe ii] 1.2570 $\mu$m/1.6442 $\mu$m, and 1.3209 $\mu$m/1.6442 $\mu$m line ratios present values of 1.79$\pm$0.20 and 0.48$\pm$0.07, both higher than the theoretical ones. This is probably owing to an inaccurate subtraction of telluric emission nearby the [Fe ii] 1.6442 $\mu$m line that makes the flux measured for this line unreliable. In Table 8 we present the results obtained from individual [Fe ii] lines in both the UVES and FIRE wavelength ranges.

5.2 Optical recombination lines

We calculate ionic abundances using He, C, N, O and Ne optical recombination lines (ORLs). The atomic data used are shown in Table 10. We detect several permitted lines of heavy-element ions, such as C ii, N ii, N iii, O i, O ii, O iii, Ne ii and Mg ii, but some are affected by fluorescence or contaminated by telluric features. We only consider pure recombination lines in these calculations. Discussion about the formation mechanism of several permitted lines can be found in Esteban et al. (1998) and Esteban et al. (2004). In Fig. 4 we show the brightest C ii and O ii lines in our optical spectrum. The high resolution of the data allow to deblend the multiple features in the region of the multiplet 1 O ii lines.

The abundance discrepancy problem arises from the fact that ORLs provide ionic abundances that are systematically larger than those obtained from CELs in photoionized nebulae, whether H ii regions or PNe. Solving this problem has critical implications for the measurement of the chemical content of nearby and distant galaxies, which is most often done using CELs from their ionized interstellar medium. Three main scenarios have been proposed to explain this discrepancy: i) the existence of temperature fluctuations over the observed volume of the nebula (Peimbert, 1967; Torres-Peimbert, Peimbert, & Daltabuit, 1980); b) cold and dense H-poor inclusions in which the bulk of the ORL emission originates (e. g. Liu et al., 2000) and, c) the departure of the free electron energy distribution from the Maxwellian distribution ($\kappa$-distribution, see Nicholls, Dopita, & Sutherland, 2012). While there is no direct observational evidence that favours or discards any of these scenarios, indirect evidence suggests that metal-rich inclusions may be the source of ADFs in PNe (Tsamis et al., 2004; Wang & Liu, 2007; García-Rojas et al., 2016).

Importantly, previous studies have shown that the ADFs of C, N, and O ions in Galactic PNe are very similar (Liu et al., 2000; Wang & Liu, 2007), and hence abundance ratios found from ORLs are suitable proxies for CEL abundance ratios. This is most relevant in the case of C/O, since no collisionally-excited lines of C ions lie in the UVES/FIRE spectral range.

NGC 5315 has a relatively low degree of ionization, and thus we do not detect He ii lines. He*+* abundances are computed with PyNeb, using the three brightest He i lines $\lambda\lambda$4471, 5876 and 6678. The effects of collisional contributions and optical depth in the triplet lines are considered. Results from the three lines are very consistent and are shown in Table 11.

C2+ abundances from ORLs are in very good agreement between lines belonging to different multiplets (see Table 11). C ii $\lambda$9903.46 is blended with the very faint [Kr iii] $\lambda$9902.30 line. Although this contribution should be negligible, we do not consider this line, because it yields a higher abundance from FIRE data than from UVES data. We ascribe this discrepancy to atmospheric absorption effects visible in the 2D optical spectrum. In general, C2+ abundances from ORLs are in very good agreement with those computed by Peimbert et al. (2004) for this object.

Fang et al. (2011) compute very detailed recombination coefficients for N ii lines. We have used these atomic data to compute N2+ abundances from multiplet 3 N ii lines. We find an excellent agreement between all the lines of the multiplet.

To compute O2+ abundances from ORLs we use lines from multiplets 1, 2, 10 and 20 as recommended by Esteban et al. (2004). Given the high density of NGC 5315, departures from LTE for O ii multiplet 1 (Ruiz et al., 2003; Tsamis et al., 2004) that are important when $n_{\rm e}$ $<$ 104 cm*-3* can be ignored. The agreement in O2+ abundances derived from the different multiplet 1 lines shows that this is a good assumption (Table 11). We use recombination coefficients assuming LS-coupling from Storey (1994) for 3s-3p transitions (multiplets 1 and 2), and the intermediate-coupling scheme by Liu et al. (1995) for 3p-3d transitions (multiplets 10 and 20). Results are shown in Table 11.

We compute Ne2+ abundances from ORLs using lines of multiplets 1, 39, 55 and 57. A detailed discussion of the excitation mechanisms of these lines can be found in García-Rojas et al. (2015). We use recombination coefficients by Kisielius et al. (1998) for multiplets 1 and 39 and those by Kisielius (private communication) for multiplets 55 and 57. We adopted the average abundance given from all these multiplets.

Finally, we estimate Mg2+ abundance from the Mg ii 3d–4f $\lambda$4481 line. For this ion there are no recombination coefficients available, so we assume that the Mg ii $\lambda$4481 line has an effective recombination coefficient equal to that of the C ii $\lambda$4267 line (Barlow et al., 2003; Wang & Liu, 2007).

6 Total abundances

To compute total abundances we have to take unobserved ions into account. In general for the most common elements we used the detailed ionization correction factors (ICFs) provided by Delgado-Inglada et al. (2014, hereinafter D-I14) to correct for the presence of unobserved ions. However, we make an exception for N (see below). In Table 12 we present the total abundances we computed for NGC 5315. The first 4 columns of Table 12 present the abundances obtained from our data from both CELs and ORLs, either by summing the ionic abundances or by applying an ICF. The next columns show the results obtained by Peimbert et al. (2004), Pottasch et al. (2002), Tsamis et al. (2003) and Tsamis et al. (2004). The last column of Table 12 shows the abundances from our optimized Cloudy model. As can be seen in Table 12, there is general good agreement within the uncertainties between our computed elemental abundances and those derived by our model and in previous works. Below we briefly discuss the most common elements, and in Section 6.1 we discuss elemental abundances of n-capture elements.

In the spectrum of NGC 5315 there are no He ii lines and ions with similar or higher IP are not seen. Given that (O2+/(O*++O2+) = 0.85, the contribution of neutral He to the total He abundance should be negligible (D-I14) and thus we assume that He/H = He+/H+*.

Similarly, we can disregard the presence of O3+ in the nebula. Thus, we computed O abundance from CELs by simply adding O*+* and O2+ abundances, which are the only two ions of this element detected in the spectrum of NGC 5315. We estimated O*+* from ORLs by scaling the ratio O*+/O2+* computed from CELs. The total O abundance from ORLs is higher, resulting in an abundance discrepancy factor, ADF(O)=1.58 (see Table 12).

Although C CELs are only detectable in UV spectra, C abundances can also be computed from optical C ii and C iii ORLs. However, the excitation of NGC 5315 is too low to detect C iii RLs. We compute the elemental C abundance using our C2+ abundance from C ii ORLs and the ICF presented in equation 39 of D-I14. We compare our results with that obtained assuming a simple sum of C*+* and C2+ abundances from ORLs; to estimate C*+/H+* we rescale using the C*+/C2+* ratios obtained from CELs by Pottasch et al. (2002) and by Dufour et al. (2015). Both results are in very good agreement with what we found using the ICF correction (see table 12). Despite the well-known abundance discrepancy problem between CEL and ORL abundances (see Sect. 5.2), numerous studies have shown that C/O (and N/O) ratios computed from ORLs are consistent with the ratios derived from CELs (e.g., Tsamis et al., 2004; Wesson, Liu, & Barlow, 2005; Wang & Liu, 2007; Delgado-Inglada & Rodríguez, 2014).

To compute the N elemental abundance from optical and NIR spectra is a delicate matter. We have computed N/O using all the available lines and different methods to better constrain this abundance ratio. First, we used the [N ii] lines to compute N+/H+. D-I14 argue that the classical scheme N/O = N+/O+ may underestimate N abundances, especially when the temperature of the central star is low, which is the case when He ii lines are not observed, and they proposed a new ICF. However, in a later work, Delgado-Inglada et al. (2015) found that their ICF could introduce an unexpected trend with the degree of ionization in the N/O values obtained for a group of H ii regions and PNe. The trend seems to be related to the fact that the nebulae are either matter- or density-bounded. The classical ICF seems to produce more accurate results for radiation-bounded nebulae, while the ICF from D-I14 is preferred for matter-bounded nebulae. To better constrain the N ICF, we compared observed ionic fractions (O2+/O*+, S2+/S+, Ar+3*/Ar2+, and He2+/He*+*) to the grid of photoionization models of D-I14. From our inspection we concluded that NGC 5315 is most likely radiation-bounded, and hence the classical ICF provides a more reliable value of the N/O in this PN. The derived N/O value from CELs and the classical ICF is 0.55. The uncertainties in log(N/O) associated with this ICF, estimated from the grid of photoionization models by D-I14, are ${}^{+0.3}_{-0.2}$ dex.

Secondly, our optimized Cloudy model (Section 3) produces a somewhat lower N abundance than the one derived with the ICF method (12+log(N/H)=8.39 compared to 8.52), and so is the N/O ratio (0.39). As a third estimate, we computed the total N/O ratio from ORLs (see Table 12). Since permitted optical N i lines are strongly affected by fluorescence effects, it is not possible to determine the N+/H+ abundance from ORLs. Therefore we use the approximation N/O = N2+/O2+ for ORLs, which gives a ratio of 0.64. The uncertainties in log(N/O) associated with this ICF are ${}^{+0.15}_{-0.05}$ and they were obtained from the D-I14 grid of photoionization models. We use the average N/O ratio weighted by the uncertainties from these estimates, 0.59${}^{0.24}_{0.15}$, in the remainder of our analysis.

We calculate total abundances of Ne, S, Cl and Ar, using the ICFs developed by D-I14 when needed, or by summing the ionic abundances when a sufficient number of ions are detected. The optical range hosts several [Ne iii], [Ne iv], and [Ne v] CELs, but owing to the relatively low ionization degree of NGC 5315 we only detect [Ne iii] lines. There are no lines of the Ne*+* ionization stage in our spectral range, and to correct for its contribution we used the ICF given by equation 17 of D-I14. We detect [S ii] and [S iii] lines in our spectra. D-I14 found that if [S ii] and [S iii] are seen in the spectra and the object does not show He ii emission, more highly-charged states have a negligible contribution. Thus, we compute the total S abundance by adding the S+ and S2+ ionic abundances. Our results are consistent with values from literature (see Table 12). Cl ions have several emission lines in the optical range, including [Cl ii], [Cl iii] and [Cl iv]. The amount of Cl4+ is negligible for NGC 5315, because the IP is too high (53.5 eV). We follow the recommendation by D-I14, who suggest that total Cl abundance can be computed as the sum of ionic abundances when [Cl ii], [Cl iii] and [Cl iv] lines are detected and He ii lines are not seen in the spectra. Finally, we detect [Ar iii] and [Ar iv] lines. Given the ionization degree of NGC 5315, we have to take into account the contribution of Ar*+, which can be an important contributor to the total abundance of Ar in low ionization PNe. The contribution of Ar4+* is negligible owing to its high IP (59.81 eV) and the non detection of He ii lines. We used the ICF shown in equation 36 by D-I14, which is valid when O2+/(O*++O2+*)$>$ 0.5. The uncertainties associated with this ICF are higher than for other elements (D-I14), but this ionization correction scheme is the best one available for this specie and our result is consistent with values from literature (see Table 12).

We assumed Mg/H=Mg2+/H+ given the wide ionization potential interval occupied by Mg2+ (Barlow et al., 2003; Wang & Liu, 2007). We obtain 12+log(Mg/H)=7.56$\pm$0.10 (see Table 12). This value is compatible with the solar photospheric value of Asplund et al. (2009) (12+log(Mg/H)=7.60) and is identical to the average value of the Mg/H ratio in Galactic disk PNe (Wang & Liu, 2007).

We used two different ICFs to calculate the total abundance of Fe, Equations 2 and 3 of Rodríguez & Rubin (2005). These two ICFs provide a range of Fe abundances. Additionally, we estimate the total Fe abundance from the sum of Fe+ and Fe2+ as 5.40$\pm$0.10. This value represents a lower limit of the total abundance of this element, since we ignore the contribution of [Fe iv], whose optical lines are intrinsically weak and were not detected. Furthermore, the contribution of Fe+ is uncertain, since its lines can arise both from the ionized portion of the nebula and from the neutral and partially molecular region. However, this sum agrees within the uncertainties with the lower limit of Fe abundance obtained from equation 3 of Rodríguez & Rubin (2005).

D-I14 do not provide corrections for unseen ions of P. In order to compute the total abundance of P, we took advantage of the similarity between the IP of P*+* and S*+* using the ionization correction scheme P/P*+=S/S+*. A detailed set of photoionization models is needed to find a more accurate ICF for this element.

6.1 Neutron-capture element abundances

The first ICF for Kr were provided by Sterling et al. (2007), through detailed photoionization models. Since then, atomic data for n-capture ion transitions have considerably improved. New atomic data were incorporated in the new ICF computations for Kr and Se by Sterling et al. (2015). They performed a large grid of photoionization models in order to cover a considerable range of physical conditions. They derived 6 ICFs for Kr and 3 for Se, depending on which ions are detected. The quality of their Kr ICFs is confirmed by optical/IR data for 16 PNe (Sterling et al., 2015; García-Rojas et al., 2015). Thanks to the high ionization degree of the PN NGC 3918, García-Rojas et al. (2015) detected several ionization stages of Kr in the optical spectrum. Therefore, they could use the complete set of ICFs provided by Sterling et al. (2015) for Kr, finding very good agreement between all the results (see first row of their Table 17). Our results for NGC 5315 are shown in Table 13. We choose the Kr abundance obtained with equation 4 of Sterling et al. (2015), because it incorporates both Kr2+ and Kr3+. Equation 1 provides a value that is very similar to the one selected. Equation 3 produces larger Kr abundances than the best empirical estimate when Ar3+/Ar $<$ 0.20 (Sterling et al., 2015), which is the case of NGC 5315. The value provided by equation 2 is consistent within the uncertainties with those derive from equation 4, but it has large error bars, owing to the uncertainty in the Ar abundance.

Selenium abundances were computed using all the ICFs proposed by Sterling et al. (2015). The ICF of equation 7 is based only on Se2+ ionic abundance, which is uncertain due to the weakness of the [Se iii] 1.0992 $\mu$m line; Equation 8 provides an ICF that only considers Se3+ and gives a value which is much lower than those provided by the other equations. Sterling et al. (2015) argue that Equation 8 is the most uncertain of the Se ICFs, due to the lack of a strong correlation with fractional abundances of light element ions. We adopt the Se abundance derived with Equation 9 of Sterling et al. (2015), which is the most accurate since it takes into account both Se2+ and Se3+. Furthermore, the unblended line of [Se iii] at 1.0992 $\mu$m, first identified by Sterling et al. (2017), allows for improved accuracy of Se abundances in PNe. The Se abundances derived with the various ICFs of Sterling et al. (2015) are shown in Table 13.

At present, there are no reliable ICFs for Xe, Br and Rb available in the literature. We calculated the total abundance of Xe using Equation 3 of Sterling et al. (2015) as the ICF, given the similarity between the IPs of Xe3+ and Kr3+. Our analysis yields a higher enrichment of Xe than for Kr (0.86 dex relative to Ar, compared to 0.02 dex for Kr), but we emphasize the uncertainty of the Xe abundance. First, Equation 3 of Sterling et al. (2015) gives systematically high Kr abundances in low- and moderate-excitation PNe (including NGC 5315) compared to other ICFs. We believe that this excitation effect plays a significant role in the anomalously large Xe abundance we derive. Secondly, the [Xe iv] $\lambda$7535.40 line is marginally detected, and [Xe iv] $\lambda$5709.21 not at all, resulting in a quite uncertain Xe3+/H+ ionic abundance.

As we discussed in Section 3.1, we possibly detect [Br iii] $\lambda$6556.56. Since the IP of Br2+ is similar to the IP of Se2+, we calculated the total abundance of Br by using equation 7 of Sterling et al. (2015). The Br abundance is shown in Table 12. The flux of the line detected at 6555.56 $\AA$ gives a Br2+ abundance much higher than expected, given the nearly solar abundances of the adjacent elements Se and Kr. We inspected the spectra to check possible contamination or alternative identifications for the feature at 6555.56 $\AA$, but we did not find any transition with other detections belonging to the same multiplet in our spectra. We also computed an upper limit to the Br2+/H+ abundance from [Br iii] $\lambda\sim$6130.40, which we did not detect. This upper limit gives a bromine abundance $<$2.43, more than an order of magnitude lower than that derived from $\lambda$6556.56. This result suggests that the [Br iii] $\lambda$6556.56 is contaminated and/or incorrectly identified in NGC 5315, and we conclude that our derived Br abundance is not reliable.

Finally, from an upper limit flux estimation of the [Rb iv] $\lambda$5759.55 line, we computed an upper limit (3$\sigma$) abundance for Rb3+. Considering the similarity with the IP range of O2+, we used the ionization correction scheme Rb/Rb3+ = O/O2+ (Sterling et al., 2016).

7 Discussion

7.1 CNO Abundances

The N/O and the C/O ratios are crucial for investigating the nature of PN progenitor stars. In particular, the C/O ratio is an indicator of the chemistry of the nebula (Sterling & Dinerstein, 2008, and references therein) and it reveals crucial information about TDU events during the thermally-pulsing AGB. The N/O ratio is a sensitive probe of HBB at the base of the convective envelope, which only occurs in stars with masses $\gtrsim 4$M*⊙*. In this section, we compare our determinations with those from the literature (see Table 14) both to assess the accuracy of our abundances as well as to utilize information from other spectral regions.

Prior to our study, the deepest optical spectrum of NGC 5315 was that analyzed by Peimbert et al. (2004). They computed N/O = 0.93 using a combination of ORLs and CELs, and C/O = 0.93 from ORLs, assuming that temperature fluctuations are responsible for the ADFs. Torres-Peimbert & Peimbert (1977) assumed the temperature fluctuations paradigm in their calculations and found N/O = 0.95 from optical observations. Similarly high N/O ratios were found by Pottasch et al. (2002) and Milingo et al. (2010). In particular, Pottasch et al. (2002) used UV observations in which N iii$]$ and N iv$]$ were detected. Along with optical data, these detections allowed them to compute the total N abundance by summing the N*+, N2+* and N3+ ionic abundances, eliminating the need for an ICF.

However, not all authors have found such large N/O ratios. Tsamis et al. (2003) utilized IUE, optical, and infrared data to compute N/O = 0.54 from CELs, although their C/O = 0.35 is lower than our derived value. This group also determined abundances from ORLs (Tsamis et al., 2004), finding N/O = 0.47 and C/O = 0.58. Dufour et al. (2015) used long-slit $HST$-STIS co-spatial UV-optical spectra and computed N/O = 0.41, which is somewhat lower than our derived value and C/O = 0.64 which is consistent with our C/O ratio; the combination of optical and UV data allowed them to avoid using ICFs. Liu et al. (2001) found N/O = 0.48 from a far-infrared ISO spectrum. Finally, de Freitas Pacheco et al. (1991) and Samland et al. (1992) derived N/O = 0.36 and N/O = 0.59, respectively, from low-resolution optical spectra.

The discrepancy seen in the N/O ratios calculated in the literature can be ascribed to various factors, including the temperature sensitivity of UV N iii$]$ and N iv$]$ emission lines, and the large and uncertain ICFs required for optical spectra in which [N ii]is the only N ion with CELs. However, the average value calculated from the literature is N/O = 0.63, which is consistent within the uncertainties with our result (N/O = 0.59${}^{+0.24}_{-0.15}$).

On the other hand C/O ratios from both ORLs and CELs seem to agree better, with the exception of the low value found by Tsamis et al. (2003) (Table 14). All extant C/O determinations are below unity, indicating that NGC 5315 has an oxygen-rich chemistry. Interestingly, this PN exhibits both O-rich crystalline silicate dust emission features and PAH features that are often associated with C-rich chemistries (Cohen & Barlow, 2005). This dual-dust chemistry has been found in other PNe with $[$WC$]$ central stars (e.g., Perea-Calderón et al., 2009) and may be associated with binary interactions (de Marco & Soker, 2002).

Oxygen-rich PNe can be associated either with low mass stars (M $<$ 1.5 M*⊙), in which TDU events did not occur (preventing C enrichment on the stellar surface), or massive stars (M $>$ 3–4 M⊙*), where the temperature at the base of the convective envelope is high enough to active HBB. In the latter case, C is converted to N by the CN cycle, leading to N and He enrichments.

Peimbert Type I PNe exhibit N and/or He enrichments typical of HBB, and hence are believed to be descendants of more massive AGB stars. This is supported by statistical evidence based on the scale heights, large central star masses and high central star temperatures of Type I PNe (e. g. Peimbert, 1990; Corradi & Schwarz, 1995; Gorny et al., 1997; Stanghellini et al., 2002; Peña et al, 2013). Several different criteria have been proposed for Type I classification in the literature. The initial criterion defined by Peimbert (1978) was overly restrictive because of the overestimation of He abundances and of electron temperatures which led to artificially large N/O ratios. Therefore, Peimbert & Torres-Peimbert (1983) refined these thresholds and defined Type I PNe as those having He/H $>$ 0.125 or N/O $>$ 0.5. Later, Maciel & Quireza (1999) required that both of these thresholds be met for Type I classification. Kingsburgh & Barlow (1994) proposed that Type I PNe be classified on the basis of the N/O ratio alone, with the minimum N/O given by (C+N)/O in H iiregions of the host galaxy. In the solar neighborhood, this limit is N/O $>$ 0.8. Based on more recent H iiregion and solar abundances, Henry, Kwitter, & Balick (2004) refined this criterion to N/O $\geq$ 0.65.

Our He/H (0.129$\pm$0.009) and N/O (0.59${}^{+0.24}_{-0.15}$) abundance determinations indicate that NGC 5315 is a Type I PN according to the Maciel & Quireza (1999) criterion, but not according to Henry, Kwitter, & Balick (2004). The uncertainties are sufficiently large that we cannot classify this PN either as a Type I or non-Type I with confidence. Literature abundances are similarly inconclusive (Table 14). Nevertheless, it appears safe to say that NGC 5315 experienced some N and He enrichments, but HBB apparently did not operate efficiently in its progenitor. This suggests that the progenitor of NGC 5315 may have had an initial mass M $\approx$ 3–5 M*⊙*. However, in this mass range the cessation of HBB can occur before the end of the thermally-pulsing AGB phase, and subsequent TDU episodes can lead to carbon enrichment (e.g., Frost et al., 1998; Ventura et al., 2015).

To shed more light on the nature of the progenitor star, we compare our abundances with the nucleosynthetic models of Karakas & Lugaro (2016, hereafter KL16), (Di Criscienzo et al.2016, hereafter DC16) and Cristallo et al. (2015, hereafter CS15). In Figure 5, we show the predicted N/O vs. He/H ratios and N/O vs. C/O ratios for different progenitor masses with solar or near-solar metallicity, compared with our observed values. In the N/O vs He/H plot the mass of the progenitor star of NGC 5315 is consistent, within the uncertainties, with predictions for 3–3.5 M*⊙* stars according to the predictions of KL16 and DC16. The CS15 models do not predict HBB at these masses at solar metallicities, and the predicted mass in this case is higher (5.5–6 M*⊙). Literature abundances generally agree with this assessment, although the Pottasch et al. (2002) and Peimbert et al. (2004) data suggest a higher initial mass of 4.5–6.0 M⊙* on account of their larger N/O ratios. On the other hand, our N/O and C/O abundances (lower panel of Figure 5) are not consistent with model predictions, except possibly for 1–1.5 M*⊙* values of KL16. Literature abundances show a dichotomy, with those having larger N/O ratios closer to the 4–5 M*⊙* predictions of KL16, and those with smaller N/O agreeing more with 4–6 M*⊙* predictions from CS15 or KL16’s 1–1.5 M*⊙* values.

Therefore light element abundances do not significantly constrain the progenitor mass of NGC 5315, other than to suggest that an initial mass of 2–3 M*⊙* is unlikely. It should be noted that while He and N are moderately enriched, as may be expected for a more massive progenitor, extra mixing processes in less massive stars during the RGB and AGB phases can also produce such enrichments (Nollett et al., 2003).

7.2 Neutron-Capture Element Abundances

The pattern of s-process enrichments of n-capture elements, or lack thereof, can also provide information about the nature of PN progenitor stars. To assess whether the s-process and TDU occurred during the AGB, it is necessary to assume an initial abundance pattern and to adopt a metallicity reference element. For Galactic disk PNe such as NGC 5315, scaled solar abundances are usually a reasonable approximation (Henry, Kwitter, & Balick, 2004). Historically, O was used as reference element in non-Type I PNe, while in Type I PNe Ar turns out to be the best choice if O is affected by nucleosynthesis in the progenitor star of these objects (Sterling & Dinerstein, 2008; Karakas et al., 2009). Moreover, Delgado-Inglada et al. (2015) have found evidence for O enrichment in Galactic PNe with carbon-rich dust, suggesting that, in general, Ar or Cl are better reference elements. A positive value of [X/(O, Ar)]22footnotetext: The notation [X/Y] represents the logarithmic difference between nebular and solar ratio abundances, i. e. [X/Y]=log(X/Y)-log(X/Y)⊙., where X is a n-capture element, indicates that the progenitor star experienced s-process enrichment. We use Asplund et al. (2009) as the reference for the Solar abundances.

In Table 15 we show the average value of Se and Kr enrichments found by Sterling et al. (2015) in their sample of Type I and non-Type I PN, compared with NGC 5315 and García-Rojas et al. (2015)’s results for NGC 3918. Given the difficulty of classifying NGC 5315 as a Type I or non-Type I PN, we decided to calculate enrichments using Ar as reference element, and we chose the same reference for the sample of García-Rojas et al. (2012); García-Rojas et al. (2015). We find that Kr and Se are not enriched, in contrast to the C-rich NGC 3918 García-Rojas et al. (2015).

Rb is another key element for diagnosing s-process nucleosynthesis, as it is an indicator of the main neutron source during the AGB phase. A large enrichment of Rb relative to other n-capture elements indicates that 22Ne is the dominant reaction for the production of neutrons. In this case, the high neutron density activates different branching points in the s-process path, increasing the production of Rb and Kr (Abia et al., 2001; van Raai et al., 2012; Karakas et al., 2012). According to models, stars with M $>$ 5 M*⊙* attain the temperatures required for the activation of the reaction 22Ne($\alpha$, n)25Mg (Busso et al., 1999), and thus Rb is an indicator of the mass of the progenitor star (see Sect. 1). We do not detect Rb in NGC 5315, but compute a 3-$\sigma$ upper limit of [Rb/Ar]$<$0.02. The lack of enrichment agrees with the Se and Kr abundances, and suggests that the progenitor of NGC 5315 is less than 6 M*⊙* (Karakas & Lugaro, 2016).

Our Br and Xe abundances (see Table 15) must be taken with caution. The identification of the $\lambda$6556 line as [Br iii]is questionable, as discussed in §6.1. In the case of Xe, we marginally detected one line, [Xe iv] $\lambda$7535.40. For lack of Xe ICFs, we use Equation 3 of Sterling et al. (2015) to compute the Xe abundance. This ICF is known to produce systematically high ICFs in low- and moderate-excitation PNe, as was found specifically for NGC 5315. For this reason, we believe that the derived Xe abundance is overestimated by $\sim$0.60 dex, based on our results for Kr (Table 13).

The lack of s-process enrichments indicate that the progenitor of NGC 5315 either was a low-mass star ($M<1.5$ M*⊙) or an intermediate-mass star (3 M${}_{\odot}<M<6$ M⊙*).

The lack of carbon enrichment is consistent with our derived n-capture element abundances. Theoretical low-to-intermediate mass stellar evolution models predict a correlation between n-capture element and C enrichment in PNe, as they are processed in the same stellar layers and dredged up together to the surface during TDU episodes (Gallino et al., 1998; Busso et al., 2001, KL14). In Figure 6 we plot the correlation between C/O and $[$Kr/Ar$]$ for PNe with the most accurate Kr abundances (derived from multiple Kr ions, see Sterling et al., 2015). We find evidence for a correlation, given by:

[TABLE]

New optical/NIR observations will be necessary to expand the sample and to strengthen this result, and search for correlations between C/O ratios and other n-capture elements.

7.3 The Enigmatic Progenitor Star of NGC 5315

Armed with our abundance analysis, we now consider the various scenarios for the nature of the progenitor of NGC 5315. The result is rather unsatisfying, as we cannot discriminate between a low-mass and intermediate-mass origin for the progenitor. The O-rich chemistry and lack of s-process enrichment only seem to rule out the mass range 1.5 M${}_{\odot}<M<3.0$M*⊙*, according to single star evolution. The conclusions are even less restrictive if binary interactions led to the formation of NGC 5315 – although there is evidence that this indeed may be the case, as outlined below.

We find three viable evolutionary channels for the progenitor of NGC 5315.

i) The intermediate-mass star scenario. NGC 5315 exhibits moderate enrichments of He and N, as may be expected for a more massive progenitor star that underwent HBB. However the N/O ratio appears to fall below the Type I PN classification criteria, indicating that HBB did not modify its composition to a substantial extent. The comparison between observed and predicted N/O and He abundances are consistent within the uncertainties with a progenitor mass of 3–6 M*⊙, depending on the AGB models used. However, in the C/O vs. N/O plane, the measured abundances are consistent with an intermediate-mass progenitor only for the model predictions of Cristallo et al. (2015). Comparison with the Karakas & Lugaro (2016) predictions instead indicate a low-mass progenitor ($M<1.5$ M⊙*). Sterling et al. (2015) found a lack of s-process enrichments of Se and Kr in Type I PNe, perhaps due to the relatively small intershell mass and/or strong dilution of enriched material into the massive envelopes of the AGB progenitor stars. The lack of s-process enrichments in NGC 5315 is consistent with this scenario.

ii) The low-mass progenitor scenario. The lack of s-process enrichments and O-rich chemistry of the nebula can be equally well ascribed to a low-mass progenitor ($M<1.5$ M*⊙* Karakas & Lugaro, 2016, and references therein) that did not experience the s-process or TDU. The He and N enrichments can be produced by extra mixing processes in low-mass stars during the RGB and the AGB phase (Nollett et al., 2003), although the nature of these phenomena are not well understood (e. g. Karakas & Lugaro, 2016). Additional support is provided by indirect estimates of the central star mass. Using a luminosity based on the extinction distance, Pottasch & Bernard-Salas (2010) estimate M$\leq$1.5 M*⊙* when comparing with post-AGB evolutionary tracks. From photoionization models, Henry et al. (2015) estimated the central star luminosity and used evolutionary tracks to determine a progenitor mass of 1.1 M*⊙. Both of these mass estimates carry caveats. The abundance analysis of Pottasch & Bernard-Salas (2010) indicated a much larger progenitor mass of M = 4.5 M⊙*. The mass derived by Henry et al. (2015) is strongly model dependent, and their stellar luminosity disagrees both with statistical distance determinations and with the extinction distance of Pottasch & Bernard-Salas (2010). Further evidence against this scenario comes from $\alpha$-element abundances. Because low-mass stars have longer lifetimes, the abundances of elements such as Cl and Ar should be lower than in Galactic H ii regions. Esteban et al. (2015) determined 12+log ($<$Cl/H$>$)=5.09 and García-Rojas & Esteban (2007) found 12+log($<$Ar/H$>$)=6.52 in H iiregions, which are lower than we find for NGC 5315. Furthermore, we compare our O, Cl and Ar abundances with the averaged values from a large sample of Type II PNe analyzed by Kwitter & Henry (2001). They found lower abundances than our results for elements unaffected by stellar nucleosynthesis: 12+log ($<$O/H$>$)=8.74, 12+log ($<$Cl/H$>$)=5.28 and 12+log ($<$Ar/H$>$)=6.45, suggesting that the progenitor star of NGC 5315 was formed more recently and hence is probably descent from a more massive progenitor star.

iii) The binary progenitor scenario. The peculiar multi-polar morphology of NGC 5315 can be produced by binary interactions during the AGB (Balick & Frank, 2002; de Marco, 2009; Hillwig et al., 2016, and references therein). Mass transfer can reduce the envelope mass of the AGB star and prematurely truncate the AGB lifetime before significant C and s-process enrichments can occur. Depending on the orbit of the binary companion, the N and He enrichments could have been produced by HBB, or by extra mixing (or some other mixing process triggered by binary interactions) (Nollett et al., 2003). This scenario is supported by the findings of Manick et al. (2015), who conducted a detailed study of radial velocity (RV) variations of a sample of central stars of PNe. They found a strong RV variation for the central star of NGC 5315, which may be explained with the presence of a close companion. Manick et al. (2015) emphasize that this result is preliminary because their observations only include four epochs, which is not sufficient to conclude that it has a periodic RV variation typical of binary systems. A further complication is the Wolf-Rayet central star itself, whose stellar wind could produce RV variations. A binary central star can also explain the dual-dust chemistry of NGC 5315 and other PNe with $[$WC$]$ central stars (de Marco & Soker, 2002; Cohen & Barlow, 2005; Perea-Calderón et al., 2009). Finally, a binary origin could help to bridge the dissonance between progenitor mass estimates from abundances and from other methods (Pottasch & Bernard-Salas, 2010).

We speculate that NGC 5315 most likely had a binary progenitor star, although we cannot dismiss the first two scenarios. The combination of RV variations, PAH and silicate dust emission features, and an overall lack of agreement between observed and predicted abundances support this scenario. The moderate N and He enrichments and Rb upper limit suggest that NGC 5315 could not have had a (single) progenitor mass above $\sim$5 M*⊙. For stars in this mass range, HBB can cease before the thermally-pulsing AGB phase is over, allowing for TDU to produce C and s-process enrichments (Frost et al., 1998; Ventura et al., 2015). The $\alpha$-element abundances of NGC 5315 suggest that a low-mass progenitor ($M<1.5$M⊙*, scenario (ii)) is unlikely.

However, more data are needed to reveal the nature of NGC 5315’s progenitor star. Specifically, observations are needed to test whether the RV variations found by Manick et al. (2015) are periodic. If the central star indeed has a binary companion, its orbital characteristics can be used to constrain the likelihood of mass transfer during the AGB. Furthermore, a detailed analysis and modeling of the $[$WC4$]$ central star (e.g. Koesterke & Hamann, 1997) can constrain the stellar luminosity and hence mass. Marcolino et al. (2007) and Todt et al. (2015)used detailed stellar atmosphere codes which are appropriate when non-LTE conditions and stellar winds are present. However, as the distance estimates to Galactic PNe are unreliable, they assumed a ‘standard” luminosity for their models and scale their results to other luminosities through a transformed radius (see Todt et al., 2015). The upcoming GAIA data release of precise parallaxes will turn around this behaviour. This information can be used to more accurately estimate the initial mass than has been possible.

These different scenarios highlight other open questions. The origin of non-spherical PNe has always been a matter of debate. There is still no quantitative theory to explain how a massive single star can form a highly non-spherical PN. The multipolar morphology can be explained with central binary systems, but only in a small fraction of them the orbital period is short enough to allow the companion to interact with the AGB star and modify the morphology of the PN (de Marco & Izzard, 2016). Therefore, the small fraction of single massive stars and compact binary systems cannot explain the high concentration of non-spherical PNe ($\approx$ 80 $\%$; Parker et al. (2006)). Another unresolved problem concerns the large percentage of Type I PNe ($\approx$ 20 $\%$; Kingsburgh & Barlow (1994)). The IMF suggests that stars with M $>$ 3 M*⊙* are only a few percent of the total population of our Galaxy. Therefore, the question is: how many Type I PNe come from high mass progenitor stars? Rb abundance determinations in Type I PNe are urgently needed to verify which Type I PNe are descendants of the Rb-rich AGB stars observed by García-Hernández et al. (2006, 2009). Finally, a crucial point for future works is to investigate whether high N and He abundances can be understood in the framework of binary interactions and/or extra mixing in low (or intermediate) mass stars.

8 Summary and conclusions

We analyzed the high-resolution (R$\sim$40000) optical spectrum obtained with UVES at the Very Large Telescope, and the medium-resolution (R$\sim$4800) near-infrared spectrum with FIRE at Magellan Baade telescope of NGC5315. We detect, identify and measure the intensities of about 700 lines in both spectra. Physical conditions were computed using intensity ratios of common ions. Ionic abundances were computed for ions which present available atomic data. Elemental abundances of common elements (O, Ar, Cl, N, S…) were performed using the more recent ICFs available when needed. The total abundances of n-capture elements were calculated for Kr, Se, Xe, Br and Rb, using sophisticated ICFs for Kr and Se and the best approximate correction scheme for Xe, Br and Rb. Finally, the enrichment of these elements was measured, using Ar as reference element, against solar abundances.

The aim of this work was to study the s-process in the progenitor of the PN NGC 5315. The relatively high concentration of N and He computed, may be associated to the occurrence of HBB and second dredge up, which are activated in AGB stars with M$>$3–4 M*⊙. Nevertheless, extra mixing events, thermohaline mixing (Eggleton et al., 2006) and magnetic buoyancy (Trippella et al., 2014) in less massive stars during first and third dredge up may also enhance the final amount of N and He. The causes of these processes are still poorly known. Furthermore, the multipolar morphology of NGC 5315 may be explained with the presence of a central binary system. Taking in account these considerations, we suggest that the most reasonable interpretation to the observed chemical abundances and for the absence of s-process enrichment in NGC 5315 is that a central binary system can be the cause of a poor or absent s-process enrichment, due to a strong interaction with a close companion that can reduce the AGB lifetime through mass transfer. However, we cannot rule out other two alternative interpretations: i) that this PN descend from a progenitor star with mass between 4–6 M⊙, where the massive convective envelope dilutes the n-capture enrichment during TDU events and the intershell region between He- and H-burning shells is smaller than in lower mass stars; ii) finally, a low mass progenitor star (M$<$1.5 M⊙*) that did not experience TDU, and the stellar surface is not enriched in s-processed material.

Our FIRE spectrum allows us to test the complete set of Se ICFs developed by Sterling et al. (2015). We consider the ICF given by their equation 9, which considers two different ions, as representative of Se total abundance. The identification of the non-blended [Se iii] $\lambda$1.0995 $\mu$m line provides a much more robust test for the Se ICFs, than the blended [Se iii] 8854 Å  line.

Finally, another important theoretical prediction was tested: the correlation between Kr and C enrichment. n-capture elements are formed in the same layers of C, both are dredged up together to the stellar surface during TDU episodes and then stellar winds and PN ejections expel them in the interstellar medium. Therefore, we expect a correlation which is shown in Figure 6.

Acknowledgments

This work is based on observations collected at the European Southern Observatory, Chile, proposal number ESO 092.D-0265(A). This work has been funded by the Spanish Ministry of Economy and Competitiveness (MINECO) under the grants AYA2011-22614 and AYA2015-65205-P. SM acknowledges support of the Instituto de Astrofísica de Canarias under a PhD fellowship. JG-R acknowledges support from Severo Ochoa Excellence Program (SEV-2015-0548) Advanced Postdoctoral Fellowship. NCS acknowledges partial support from NSF award AST-0901432. GD-I gratefully acknowledges support from: Conacyt grant no. CB-2014/241732 and PAPIIT (DGAPA-UNAM) grants no. 107215 and IA-101517. SM acknowledges the University of West Georgia, where part of this work was done. We thank D. A. García-Hernández for fruitful discussions. AMD acknowledges support from the FONDECYT project 3140383. We thank the anonymous referee for a detailed report that help to improve the scientific content of the paper.