3 Spectroscopic observations and results

In total, 169 candidates were observed spectroscopically. During two runs in October 1998 and May 1999, 157 objects were observed at the ESO 3.6 m, La Silla, Chile using the EFOSC spectrograph. For both runs the detector used was "ccd40'', a Loral CCD with $2048\times2048$ pixels, binned $2\times2$. For the 1999 May observations grism #12 was used, covering the spectral range 5800 Å-9500 Å with a dispersion of 4.23 Å per binned pixel. For the 1998 October observations grisms #4 and #5 were used covering the ranges 3420-7460 Å and 6000-10 000 Å respectively with dispersions of 4.08 and 4.27 Å per binned pixel. The slit width was chosen to match the seeing, typically 1.2-1.5 arcsec. For the 1998 October observations a sky PA of $270^\circ$ was used, and for the 1999 May observations the slit was at the paralactic angle. Approximate relative flux calibration was obtained using observations of spectrophotometric standards from Hamuy et al. (1994). Twelve sources were observed at AAT using the RGO spectrograph in October 1996.

26 sources remain to be observed, of which 24 are in the North Galactic Cap (NGC) and 2 are in the South Galactic Cap (SGC) region.

 

 

Table 1: Summary of current survey status for sources with $S\ge 50~\rm mJy$.

Number of candidates		Number confirmed
with z data		3.5<z<4.0	4.0<z<4.5	z>4.5
Present observations	169	7	4	2
Literature data	33	2	1	0
Total	202	9	5	2

The data were reduced using standard procedures with the IRAF software environment.

Table 1 summarises the current survey status and in Table 2 we give redshifts, optical magnitudes and radio flux densities for the objects which were identified as new high-redshift quasars. The spectra are shown in Fig. 5, and finding charts, with J2000 coordinates, are given in Fig. 7. Two of the z>4 QSOs are discussed in more detail below.

PMNJ1451-1512:

this is now the most distant radio-selected quasar. The object is blank on the UKST B-band plate, detected in R, and very bright in I(17.3). Near-IR magnitudes from the 2MASS survey are $J = 16.39\pm0.10$, $H = 15.28\pm 0.09$, $K = 14.67\pm 0.09$.

The optical spectrum (Fig. 5) shows strong lines of $\rm Ly\alpha$ SiIV/OIV and CIV from which we estimate a redshift of 4.763. However the absorption of the CIV and Ly-$\alpha$ line make redshift determination uncertain. A 4800 s IR spectrum was obtained using SOFI at the 3.5 m NTT, La Silla on 31 May 1999. The spectrum shows strong, broad emission lines of CIII and MgII at a redshift of 4.764 and approximately 4.78 respectively (the MgII line has a complex shape, see Fig. 6). We adopt a redshift of 4.763 for this QSO.

The NVSS map (see Fig. 8) shows a source at the optical position of the QSO, and a second NVSS source about 4 arcmin away. The PMN position appears to lie between the two, although much closer to the first position (that of the QSO). It is possible that the PMN flux density is overestimated because of a contribution from the second source.

Figure 4: Colour-magnitude and colour-colour plots for the APM identifications within 3'' of the NVSS radio position. Solid symbols represent objects classified as stellar, open symbols represent objects classified as galaxies by the APM. Arrows are used to represent limits on the colours when objects are not detected in one or more of the three passbands.

Figure 5: Optical spectra of the z>3 quasars.

Figure 5: continued.

PMN J0525-3343:

this QSO was discovered in the AAT run of October 1996. It was reobserved at the Las Campanas 100'' du Pont telescope using the ModSpec spectrograph and this spectrum is shown Fig. 5. The somewhat lower signal-to-noise combined with the lack of clean strong lines make the redshift determination difficult. However Péroux et al. (2001) have since obtained a higher quality spectrum from which a more accurate redshift measurement can be made. They obtain $z=4.383\pm 0.034$ assuming rest wavelengths for the emission lines that have not been corrected for systematic shifts, or z=4.338when assuming the rest wavelengths used in Table 3, which have been corrected for systematic shifts as described by Tytler & Fan (1992).

This source has been confirmed as an X-ray source with ROSAT HRI and its X-ray properties are the subject of another paper (Fabian et al. 2001).

   

Table 2: Summary of optical and radio properties for quasars with z>3 in the sample. A "-'' indicates that no data were available. The positions are the APM optical positions determined from the R plates.

Name	RA	DEC	z	R	B	I	$S_{\rm 5~GHz}$	$\rm\alpha_{5~GHz}^{1.4~GHz}$	Place & Date	comments
	J2000								discovered
PMN J0022-0759	00 22 00.246	-07 59 16.03	3.896	19.47	21.26	>19.25	54	-0.18	ESO 10/98
PMN J0214-0518	02 14 29.295	-05 17 44.55	3.986	18.42	20.54	18.75	93	0.61	ESO 11/00
PMN J0235-1805	02 34 55.143	-18 06 08.49	4.314	18.79	22.00	-	49	-0.16	ESO 11/00	$S_{5~\rm GHz} < 50$ mJy (confused; see NVSS)
PMN J0324-2918	03 24 44.280	-29 18 21.10	4.630	18.66	22.18	-	354	0.30	ESO 11/00
PMN J0326-3253	03 27 00.407	-32 54 18.65	3.463	19.40	21.15	-	78	-0.01	ESO 11/00
PMN J0525-3343	05 25 06.166	-33 43 05.34	4.413	18.50	21.47	-	210	0.06	AAT 10/96	X-ray source, Fabian et al. (2001)
PMN J1043-2140	10 43 11.963	-21 40 47.97	3.774	20.66	>22.39	>18.99	59	-0.07	ESO 05/99
PMN J1108-1804	11 08 48.035	-18 04 50.85	3.433	20.00	21.48	19.18	48	-0.31	ESO 05/99	$S_{5~\rm GHz} < 50$ mJy
PMN J1429-1616	14 29 31.410	-16 15 40.44	3.842	19.76	21.26	19.42	50	0.34	ESO 05/99
PMN J1451-1512	14 51 47.052	-15 12 19.99	4.763	19.14	>22.60	17.25	90	0.89	ESO 05/99	$S_{5~\rm GHz}$ confused; see NVSS
PMN J2042-2223	20 42 57.278	-22 23 26.69	3.630	19.70	21.52	-	184	0.23	ESO 10/98
PMN J2134-0419	21 34 12.006	-04 19 09.87	4.346	19.98	>22.80	19.73	221	-0.23	ESO 10/98
PMN J2219-2719	22 19 35.304	-27 19 02.76	3.634	19.72	21.61	-	221	-0.27	ESO 11/00
PMN J2220-3336	22 20 26.957	-33 36 59.44	3.691	21.08	22.73	-	123	0.67	ESO 11/00	$S_{5~\rm GHz}$ confused; see NVSS
PMN J2314+0201	23 14 48.722	+02 01 50.86	4.110	19.87	>22.18	>19.28	97	-0.22	ESO 10/98	X-ray source, Boller et al. (1997)
Previously known QSOs contained in the sample										Reference
PMN J1028-0844			4.276	18.84	20.72	18.37	159	-0.43		Zickgraf et al. (1997), X-ray source 0.3' away
PMN J1230-1139			3.528	19.45	21.95	-	374	-0.19		Drinkwater et al. (1997)
PMN J2003-3251			3.78	17.60	19.31	-	1248	0.77		Peterson et al. (1982)

   

Table 3: Wavelengths of emission lines and redshifts for new z>3objects, measured from the discovery spectra shown in Fig. 5. A * indicates a line affected by absorption. Redshifts are calculated assuming rest wavelengths corrected for systematic shifts (Tytler & Fan 1992) as follows: Ly-$\alpha$1214.97 Å; CIV 1547.46 Å; CIII 1906.53 Å; SiIV/OIV 1398.62 Å. Uncertainties in the redshifts are approximately $\pm$0.005, based on the range of redshift determined from different emission lines in the same QSO.

Name	Ly$\alpha$ peak	SiIV/OIV	CIV	CIII	mean z
PMN J0022-0759	5956.46	6831.46*	7583.92*	-	3.896
PMN J0214-0518	6053.03	6978.49	7714.46	-	3.986
PMN J0235-1805	6455.15	-	8225.35	-	4.314
PMN J0324-2918	6846.70	-	8703.26	-	4.630
PMN J0326-3253	5422.49	6234.08	6913.65	8506.93	3.463
PMN J0525-3343	6627.50	7536.03	8347.96	-	4.413$\dagger$
PMN J1043-2140	-	-	7388.09	9101.68	3.774
PMN J1108-1804	-	-	6863.86	8446.82	3.433
PMN J1429-1616	-	6770.20	7498.27	9229.22	3.842
PMN J1451-1512	7004.19	8057.75	8889.01*	10989.2 $\dagger\dagger$	4.763
PMN J2042-2223	5625.07	-	7166.39	-	3.630
PMN J2134-0419	6488.28	7488.82	8269.41	-	4.346
PMN J2219-2719	5628.22	6482.09	7138.42*	-	3.634
PMN J2220-3336	5684.32	6570.76	7265.08	-	3.691
PMN J2314+0201	6195.83	7153.88	7914.37	-	4.110

$\dagger$ Péroux et al. measure z=4.388 for this QSO (when using the above rest wavelengths) based on a higher-quality spectrum.
$\dagger\dagger$ Line measured from IR spectrum.