Abstract
We show that if a bounded pseudoconvex domain satisfies the solvability of the bounded \({\overline{\partial }}\) problem, then the ideal of bounded holomorphic functions vanishing at a point \(\alpha \) in the domain is generated by \((z-\alpha )\). We also prove a smooth analog of the main result for bounded pseudoconvex domains with a sufficiently smooth boundary and also consider the Bergman space case.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Preliminaries
Let \(\Omega \subset {\mathbb {C}}^n\) be a domain and we denote the space of bounded holomorphic functions on \(\Omega \) as \(H^{\infty }(\Omega )\). Here, we equip \(H^{\infty }(\Omega )\) with the usual compact open topology. We say a bounded pseudoconvex domain \(\Omega \subset {\mathbb {C}}^n\) is Gleason solvable if the ideal of functions in \(H^{\infty }(\Omega )\) vanishing at a point is finitely generated. More precisely, we have the following definition.
Definition 1
We say \(\Omega \) Gleason solvable if for any \(\alpha \in \Omega \), the ideal in \(H^{\infty }(\Omega )\) of functions vanishing at \(\alpha \) is generated by \((z-\alpha )\). That is, if \(g\in H^{\infty }(\Omega )\) and \(g(\alpha )=0\), then there exists \(g_1,\ldots ,g_n\in H^{\infty }(\Omega )\) so that \(g\equiv \sum _{j=1}^n (z_j-\alpha _j)g_j\) where \(\alpha :=(\alpha _1,\ldots ,\alpha _n)\).
As we will show, bounded pseudoconvex domains for which the \(\overline{\partial }\)-problem is solvable in \(L^{\infty }\), which we will call \(L^{\infty }\)-pseudoconvex, have the nice algebraic property of Gleason solvablility. As a reference for such domains and more generally Stein manifolds with trivial \(L^{\infty }\) \(\overline{\partial }\)-cohomology, see the classical work [5]. We will define \(L^{\infty }\)-pseudoconvex domains in a more precise manner.
Definition 2
Let \(\Omega \subset {\mathbb {C}}^n\) be a bounded pseudoconvex domain. We say \(\Omega \) is \(L^{\infty }\)-pseudoconvex if for any \(\overline{\partial }\)-closed bounded form \(\beta \in L^{\infty }_{(0,s)}(\Omega )\), there exists \(\alpha \in L^{\infty }_{(0,s-1)}(\Omega )\cap \mathbf{dom} (\overline{\partial })\) so that
Gleason solvability is sometimes referred to as Hefer’s condition (as developed by Hefer in 1940) for the ring of bounded holomorphic functions on \(\Omega \). For more information on Hefer’s condition, see [8, Chapter 5]. Many of the methods used in the study of Hefer’s condition involve the integral representation techniques developed in [4, 11]. Gleason solvability was first considered for the ball algebra on the unit ball in \({\mathbb {C}}^n\). This was first proved by Leibenson, albeit informally. See [14]. Gleason solvability for holomorphic Bergman spaces, holomorphic mixed norm spaces, and the holomorphic Bloch spaces were studied in [9, 10, 12, 14, 17]. The domains considered were intitially the unit ball, but then were generalized to strongly pseudoconvex domains with \(C^2\)-smooth boundary. There is also considerable interest on other function spaces and domains, including the weighted \(L^p\) spaces on egg shaped domains, as seen in [13]. See [6] for some results on Gleason solvability on the harmonic Bloch spaces of bounded strongly pseudoconvex domains with smooth boundary.
An application of Gleason solvable domains is a several variables partial generalization of the following commuting Toeplitz operator theorem seen in [1]. We state the theorem for the convenience of the reader.
Theorem 1
[1] Let \(\Omega \subset {\mathbb {C}}\) be a bounded domain. Suppose \(\phi \in H^{\infty }(\Omega )\) and \(\psi \) be a bounded measurable function so that \(T_{\phi }\) and \(T_{\psi }\) commute on the Bergman space \(A^2(\Omega )\). Then \(\psi \) is holomorphic on \(\Omega \).
The commuting Toeplitz operator problem in several variables remains unsolved on the Bergman spaces of bounded pseudoconvex domains. However, some partial results were obtained on the the ball, strongly pseudoconvex domains in general, and on domains where the \(\overline{\partial }\)-problem can be solved in \(L^{\infty }\). See [15] for results concerning the commuting Toeplitz operator problem on the Bergman space of such domains. In the paper [2] it was shown that the Gleason solvability condition implies \(\Omega \) can be identified with an open subset of its maximal ideal space. Furthermore, this condition implies the density of an algebra generated by conjugate holomorphic and holomorphic functions in various function spaces. Such density results are required in the proof of the commuting Toeplitz operator problem in several variables as seen in [15]. As an example, any bounded strongly pseudoconvex domain in \({\mathbb {C}}^2\) with a boundary of class \(C^4\) is Gleason solvable in the continuous case.
Theorem 2
[7] Let \(\Omega \subset {\mathbb {C}}^2\) be bounded and strongly pseudoconvex with a \(C^4\) smooth boundary. Then \(\Omega \) is Gleason solvable in the continuous case. That is, assume \(\phi \in H^{\infty }(\Omega )\cap C(\overline{\Omega })\) so that \(\phi (\alpha )=0\) for some \(\alpha :=(\alpha _1, \alpha _2)\in \Omega \). Then \(\phi \equiv (z_1-\alpha _1)\phi _1+(z_2-\alpha _2)\phi _2\) where \(\phi _1, \phi _2\) are in the same space as \(\phi \).
We will now define some terms related to the \(\overline{\partial }\)-Koszul complex. We let \(\Omega \subset {\mathbb {C}}^n \) be a domain and we define \({\mathcal {C}}_{(0,s)}:=L^{\infty }(\Omega )\cap C^{\infty }_{(0,s)}(\Omega )\) where \(C^{\infty }_{(0,s)}(\Omega )\) are (0, s)-forms that are smooth on \(\Omega \). For V an m-dimensional vector space with basis \(\{e_1,\ldots ,e_m\}\), we define \(\wedge ^r V:=\mathbf{span} \{e_{j_1}\wedge e_{j_2}\wedge \cdots \wedge e_{j_r}: j_1< \cdots <j_r\}\). Then we define the tensor product
where r and s are non-negative integers. Then we define the densely defined unbounded operator
as
We note that a similar definition holds for the \(L^2\) and \(C^{\infty }(\overline{\Omega })\) setting (where \(L^{\infty }\) is replaced with \(L^2\) in these definitions). We will denote the \(L^2\) analog of \(\Gamma ^{\infty }_{(r,s)}\) as \(\Gamma ^2_{(r,s)}\) and the \(C^{\infty }(\overline{\Omega })\) analog as \(\Gamma ^{C^{\infty }(\overline{\Omega })}_{(r,s)}\). Next, for any bounded holomorphic mapping \(F:=(f_1,\ldots ,f_n):\Omega \rightarrow {\mathbb {C}}^n\) we will define the operator
as
-
(1)
\(\tau _F(e_j\otimes W)=f_jW\)
-
(2)
\(\tau _F\overline{\partial }=\overline{\partial }\tau _F\) on \(\{f\in \Gamma ^{\infty }_{(r,s)}: \overline{\partial }f\in \Gamma ^{\infty }_{(r,s+1)}\}\)
-
(3)
\(\tau _F\tau _F=0\) and \(\overline{\partial }^2=0\).
-
(4)
\(\tau _F(A\wedge B)=\tau _F(A)\wedge B+(-1)^{|A|}A\wedge \tau _F(B)\) where |A| is the order of A in \(\bigcup _{r=0}^m \bigwedge ^r V\).
Introducing more notation, we define \({\mathcal {D}}_s:={\mathcal {C}}_{(0,s)}\cap \mathbf{dom} (\overline{\partial })\).
2 Main Results and Proofs
One can modify the proofs of [15, Lemma 1] and [15, Lemma 3] to get the following proposition, which we prove for the convenience of the reader.
Proposition 1
Let \(\Omega \subset {\mathbb {C}}^n\) be a bounded \(L^{\infty }\)-pseudoconvex domain and \(F:=(f_1,\ldots ,f_n):\Omega \rightarrow \Omega \) be injective, holomorphic, and \(C^{\infty }\)-smooth up to \(\overline{\Omega }\). Let \(W\in \Gamma ^{\infty }_{(r,s)}\cap \mathbf{dom} (\overline{\partial })\) so that
-
(1)
\(\overline{\partial }W=0\).
-
(2)
W is supported away from \(F^{-1}(0,0,\ldots ,0)\).
-
(3)
\(\tau _F(W)=0\).
Then there exists \(Y\in \Gamma ^{\infty }_{(r+1,s-1)}\cap \mathbf{dom} (\overline{\partial })\) so that \(\tau _F\overline{\partial }Y=W\).
Proof
Let \(\chi \in C^{\infty }(\overline{\Omega })\) so that \(\chi \equiv 1\) on the support of W and \(\text {supp}(\chi )\cap F^{-1}(0,0,\ldots ,0)=\emptyset \). Now define
and define
Now, \(\tau _F(X)=1\) on the support of W and so we define \(Y:=X\wedge W\in \Gamma ^{\infty }_{(r+1,s)}\). Then
and the support of Y is away from \(F^{-1}(0,0,\ldots ,0)\). Furthermore, \(\overline{\partial }Y\) is bounded since \(\overline{\partial }W=0\) and \(g_j\in C^{\infty }(\overline{\Omega })\) for \(j\in \{1,\ldots ,n\}\). Note that in [15, Proof of Lemma 3] we needed X to have compact support. This is not needed in this proposition since we assumed F is smooth up to the closure of the domain (that is, \(\{f_1,\ldots ,f_n\}\subset C^{\infty }(\overline{\Omega })\)) and injective (so \(F^{-1}(0,0,\ldots ,0)\) is a singleton set). The rest of the proof follows the descending induction procedure seen in [15, Proof of Lemma 3] with some minor modifications. Let \(s=n\), \(0\le r\le m-1\) and also \(W\in \Gamma ^{\infty }_{(r,n)}\) so that \(\text {supp}(W)\cap F^{-1}(0)=\emptyset \) and \(\tau _F(W)=0\). It is clear that \(\overline{\partial }W=0\) since any (0, n) form is \(\overline{\partial }\)-closed. Then there is \(Y_1\in \Gamma ^{\infty }_{(r+1,n)}\) so that \(\tau _F(Y_1)=W\). Also, it is clear that \(\overline{\partial }Y_1=0\). Since \(\Omega \) is \(L^{\infty }\)-pseudoconvex, there exists \(Y\in \Gamma ^{\infty }_{(r+1, n-1)}\) so that \(\overline{\partial }Y=Y_1\). In other words, \(\tau _F\overline{\partial }(Y)=W\). This is our base case. Now assume our proposition is true for \(s=k+1, k+2,\ldots ,n\) and \(r=0,1,\ldots ,m-1\). Suppose \(0\le r\le m-1\) and suppose \(W\in \Gamma _{(r,k)}^{\infty }\) has the following properties:
-
(1)
\(\text {supp}(W)\cap F^{-1}(0,0,\ldots ,0)=\emptyset \),
-
(2)
\(\overline{\partial }W=0\) and \(\tau _F(W)=0\).
Then there exists \(Y_1\in \Gamma ^{\infty }_{(r+1,k)}\) so that
-
(1)
\(\overline{\partial }Y_1\in \Gamma ^{\infty }_{(r+1,k+1)}\),
-
(2)
\(\text {supp}(Y_1)\cap F^{-1}(0,0,\ldots ,0)=\emptyset \)
Therefore, \(\overline{\partial }Y_1\in \Gamma ^{\infty }_{(r+1,k+1)}\) satisfies the conditions of the proposition for \(s=k+1\). In other words,
-
(1)
\(\text {supp}(\overline{\partial }Y_1)\cap F^{-1}(0,0,\ldots ,0)=\emptyset \),
-
(2)
\(\overline{\partial }\overline{\partial }Y_1=0\) and \(\tau _F\overline{\partial }(Y_1)=\overline{\partial }W=0\).
By the induction hypothesis, we have \(\overline{\partial }Y_2\in \Gamma ^{\infty }_{(r+2,k+1)}\) and \(\tau _F\overline{\partial }Y_2=\overline{\partial }Y_1\) for some \(Y_2\in \Gamma ^{\infty }_{(r+2,k)}\). Thus we have
Let us define \(Y_3:=Y_1-\tau _FY_2\). Then we have
and
Since \(\Omega \) is \(L^{\infty }\)-pseudoconvex, there exists \(Y\in \Gamma ^{\infty }_{(r+1, k-1)}\cap \mathbf{dom} (\overline{\partial })\) so that \(\overline{\partial }Y=Y_3\). Thus we have
as desired. \(\square \)
Remark 1
The same procedure is applicable to the space \(\Gamma ^2_{(r,s)}\) where the proof of Proposition 1 is modified slightly to accomodate the different space. In this case, we do not need a bounded \(L^{\infty }\)-pseudoconvex domain, as the \(\overline{\partial }\)-problem is solvable on \(L^2(\Omega )\) for any bounded pseudoconvex domain. See [16] for more information on the \(L^2\)-theory of the \(\overline{\partial }\)-Neumann problem.
The proof for the following Proposition is the same as for Proposition 1, except one must use the following theorem in place of the solvability of \(\overline{\partial }\) in \(L^{\infty }\).
Theorem 3
[3] Let \(\Omega \subset {\mathbb {C}}^n\) be a bounded pseudoconvex domain with a \(C^{\infty }\)-smooth boundary. Suppose \(\beta \in C^{\infty }_{(0,s)}(\overline{\Omega }) \) so that \(\overline{\partial }\beta =0\). Then there exists \(\alpha \in C^{\infty }_{(0,s-1)}(\overline{\Omega }) \) so that \(\overline{\partial }\alpha =\beta \).
Proposition 2
Let \(\Omega \subset {\mathbb {C}}^n\) be a bounded pseudoconvex domain with a \(C^{\infty }\)-smooth boundary and \(F:=(f_1,\ldots ,f_n):\Omega \rightarrow \Omega \) be injective, holomorphic, and \(C^{\infty }\)-smooth up to \(\overline{\Omega }\). Let \(W\in \Gamma ^{C^{\infty }(\overline{\Omega })}_{(r,s)}\) so that
-
(1)
\(\overline{\partial }W=0\).
-
(2)
W is supported away from \(F^{-1}(0,0,\ldots ,0)\).
-
(3)
\(\tau _F(W)=0\).
Then there exists \(Y\in \Gamma ^{C^{\infty }(\overline{\Omega })}_{(r+1,s-1)}\) so that \(\tau _F\overline{\partial }Y=W\).
The following is the main theorem. As a consequence of this main theorem, we have that the commuting Toeplitz operator problem as seen in [15] is true on a possibly larger class of bounded pseudoconvex domains.
Theorem 4
Let \(\Omega \) be a bounded \(L^{\infty }\)-pseudoconvex domain in \({\mathbb {C}}^n\). Then \(\Omega \) is Gleason solvable.
Proof
This proof is an application of a few results concerning the \(\overline{\partial }\)-Koszul complex. See [15] for more information on the \(\overline{\partial }\)-Koszul complex. We will first consider \(n=2\). Let \(\Omega \subset {\mathbb {C}}^2\) be a bounded \(L^{\infty }\)-pseudoconvex domain. Let \(g\in H^{\infty }(\Omega )\) and without loss of generality, suppose \(g(0,0)=0\). Since g has a power series expansion about (0, 0) converging on some open set \(U\subset \subset \Omega \), one can write
on U. Here, \(\lambda _1, \lambda _2\in H^{\infty }(U)\cap C^{\infty }({\overline{U}})\). Now let \(\chi _1\in C^{\infty }(U)\), supported in U, \(0\le \chi _1\le 1\) on U and \(\chi \equiv 1\) on \({\widetilde{U}}\subset \subset U\). Also, \((0,0)\in {\widetilde{U}}\). Then on \(\Omega \setminus \overline{{\widetilde{U}}}\) we have
Therefore,
on \(\Omega \). Now for the sake of notation, define
and
Notice that both \(L_1\) and \(L_2\) are bounded on \(\Omega \) and partial derivatives of \(L_1\) and \(L_2\) of all orders exist and are continuous on compact subsets of \(\Omega \). In addition, \(\overline{\partial }L_1\) and \(\overline{\partial }L_2\) are bounded on \(\Omega \). Now define
Notice that \(W_1\) satisfies the hypothesis of Proposition 1 since the support of \(W_1\) is away from \(F^{-1}(0,0)\) where \(F:=(z_1,z_2)\), \(W_1\in \Gamma ^{\infty }_{(1,1)}\cap \mathbf{dom} (\overline{\partial })\), \(\overline{\partial }W_1=0\), and \(\tau _F W_1=0\). Thus by Proposition 1, there exists there exists \(H_1\in {\mathcal {C}}_{(0,0)}^{\infty }\cap \mathbf{dom} (\overline{\partial })\) so that
and
Thus we have that
and
Therefore,
and
Furthermore,
on \(\Omega \). This completes the proof for \(n=2\). We will demonstrate the proof works for \(n=3\) and then show how it can be generalized. For \(\Omega \subset {\mathbb {C}}^n\) for \(n=3\), we first use a power series argument similar to the \(n=2\) case. That is, if \(g(0)=0\), one can write \(g\equiv z_1L_1+z_2L_2+z_3L_3\) where \(L_j\) are holomorphic on a neighborhood of 0, \(L_j\) are smooth and bounded on \(\Omega \), and \(\overline{\partial }L_j\in L^{\infty }_{(0,1)}(\Omega )\) for \(j=1,2,3\). We define
Clearly, \(\overline{\partial }W=0\) and \(\tau _F(W)=0\) where \(F=(z_1,z_2,z_3)\). Furthermore, the support of W is away from \(F^{-1}(0,0,0)\). Therefore, by Proposition 1, there exists \(Y\in \Gamma ^{\infty }_{(2,0)}\) so that
Now, \(Y\in \Gamma ^{\infty }_{(2,0)}\), therefore, Y has the form
for \(\{y_1, y_2, y_3\}\subset {\mathcal {C}}_{(0,0)}^{\infty }\cap \mathbf{dom} (\overline{\partial })\). Now we apply \(\tau _F\overline{\partial }\) to Y and collect terms together. We get
Thus we have
and
This implies
and
Furthermore, we can write
The proof for \(n>3\) is an analog of the previous proofs with more terms involved. We write g as a linear combination of bounded functions in the domain of \(\overline{\partial }\) that are holomorphic on a neighborhood of 0. Then we use Proposition 1 to ’correct’ these bounded functions to be holomorphic and bounded on \(\Omega \). Furthermore, the linear combination of these functions still gives g. \(\square \)
The advantage of this approach is that one can modify the proof to consider the Bergman space \(A^2(\Omega )\). Recall the Bergman space of \(\Omega \) is the space of square integrable holomorphic functions on \(\Omega \). The proof of this following theorem follows the proof of Theorem 4 where Proposition 1 is used in addition to Remark 1.
Theorem 5
Let \(\Omega \subset {\mathbb {C}}^n\) be a bounded pseudoconvex domain. Suppose \(\phi \in A^2(\Omega )\) and \(\phi (\alpha )=0\) for some \(\alpha \in \Omega \). Then, there exists \(\phi _1,\phi _2,\ldots ,\phi _n\in A^2(\Omega )\) so that \(\phi \equiv \sum _{j=1}^n(z_j-\alpha _j)\phi _j\).
It is well known that the boundary smoothness of the domain plays a role in determining the regularity of the solutions of the \(\overline{\partial }\)-problem with smooth data. Therefore, one can prove the following result for the \(\overline{\partial }\) problem with \(C^{\infty }\)-smooth data using the techniques developed for the \(L^{\infty }\) case (see the proof of Theorem 4) together with Propostion 2. This result is a variation of one of the main results in [7]. The proof follows the proof of Theorem 4 with Proposition 2 replacing Proposition 1.
Theorem 6
Let \(\Omega \subset {\mathbb {C}}^n\) be a \(C^{\infty }\)-smooth bounded pseudoconvex domain. Then \(\Omega \) has solvable \(C^{\infty }\)-smooth Gleason problem. That is, for any \(\phi \in C^{\infty }(\overline{\Omega })\cap A^2(\Omega )\) vanishing at \(\alpha :=(\alpha _1,\ldots ,\alpha _n)\in \Omega \), there exists
so that
Data Availability
Data sharing not applicable to this article as no datasets were generated or analysed during the current study.
References
Axler, S., Čučković, Ž, Rao, N.V.: Commutants of analytic Toeplitz operators on the Bergman space. Proc. Am. Math. Soc. 128(7), 1951–1953 (2000)
Clos, T.G., Izzo, A.J.: Approximation by an algebra generated by holomorphic and conjugate holomorphic functions. Rocky Mt. J. Math. (to appear). arXiv:2107.08332v1
Chen, S.-C., Shaw, M.-C.: Partial Differential Equations in Several Complex Variables, AMS/IP Studies in Advanced Mathematics, vol. 19. American Mathematical Society, Providence (2001)
Henkin, G.M.: Integral representations of functions holomorphic in strictly pseudo-convex domains and some applications. Math. USSR Sb. 7, 597–616 (1970) (English)
Henkin, G.M., Leiterer, J.: Theory of Functions on Complex Manifolds, Math. Lehrbücher Monogr., II. Abt., Math. Monogr., vol. 60. Akademie-Verlag, Berlin (1984) (English)
Hu, Z.: Gleason’s problem for harmonic mixed norm and Bloch spaces in convex domains. Math. Nachr. 279(1–2), 164–178 (2006)
Kerzman, N., Nagel, A.: Finitely generated ideals in certain function algebras. J. Funct. Anal. 7, 212–215 (1971)
Krantz, S.G.: Function Theory of Several Complex Variables. AMS Chelsea Publishing, Providence (2001), Reprint of the 1992 edition
Liu, Y.: Boundedness of the Bergman type operators on mixed norm spaces. Proc. Am. Math. Soc. 130(8), 2363–2367 (2002)
Ortega, J.M.: The Gleason problem in Bergman-Sobolev spaces. Complex Variables Theory Appl. 20(1–4), 157–170 (1992)
Ramirez de Arellano, E.: Ein Divisionsproblem und Randintegraldarstellungen in der komplexen Analysis. Math. Ann. 184, 172–187 (1970) (German)
Ren, G., Shi, J.: Bergman type operator on mixed norm spaces with applications. Chin. Ann. Math. Ser. B 18(3), 265–276 (1997), A Chinese summary appears in Chinese Ann. Math. Ser. A 18(4), 527 (1997)
Ren, G., Shi, J.: Gleason’s problem in weighted Bergman space on egg domains. Sci. China Ser. A 41(3), 225–231 (1998)
Rudin, W.: Function Theory in the Unit Ball of \({{\mathbb{C}}}^n\), Classics in Mathematics. Springer, Berlin (2008), Reprint of the 1980 edition
Şahutoğlu, S., Tikaradze, A.: On a theorem of Bishop and commutants of Toeplitz operators in \({{\mathbb{C}}}^n\). Rend. Circ. Mat. Palermo (2) 68(2), 237–246 (2019)
Straube, E.J.: Lectures on the \(L^2\)-Sobolev Theory of the \({\overline{\partial }}\)-Neumann Problem, ESI Lectures in Mathematics and Physics. European Mathematical Society (EMS), Zürich (2010)
Zhu, K.H.: The Bergman spaces, the Bloch space, and Gleason’s problem. Trans. Am. Math. Soc. 309(1), 253–268 (1988)
Acknowledgements
I wish to thank Akaki Tikaradze and Alexander Izzo for useful conversations and comments on a preliminary version of this manuscript. I also thank the anonymous referees for their suggestions.
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Mihai Putinar.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This article is part of the topical collection “Higher Dimensional Geometric Function Theory and Hypercomplex Analysis” edited by Irene Sabadini, Michael Shapiro and Daniele Struppa.
Rights and permissions
About this article
Cite this article
Clos, T.G. Solvability of the Gleason Problem on a Class of Bounded Pseudoconvex Domains. Complex Anal. Oper. Theory 16, 58 (2022). https://doi.org/10.1007/s11785-022-01236-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11785-022-01236-5